Skip to main content

Full text of "The palaeography of Greek papyri. By Frederic G. Kenyon ... With twenty facsimiles and a table of alphabets"

See other formats



















4. B. 43 









The substance of the following treatise was originally 
written as a dissertation for the Conington Prize in 1897, 
and I had at first intended to defer its publication for some 
time, in view of the accessions of new material which the 
excavations conducted each year in Egypt are continually 
bringing to light. It was represented, however (and that 
by one whose judgement I was bound to respect), that, 
although the work might gain in absolute accuracy by 
such a delay, yet its usefulness to students would be greater 
if it were published now, as enabling them to assimilate 
the new material for themselves. Consequently it is with- 
out any idea of having reached finality, but rather as an 
assistance to investigation, that this book is now ofiered to 
the consideration of students of palaeography. 

It is, in fact, an essay in the strict sense of the term — an 
attempt to formularize and classify the results of a number 
of discoveries, most of which have occurred quite recently. 
Hence, it will be understood, the conclusions stated in it are 
in many instances only the impressions of a single student 
of the subject, and do not (as is the case with treatises on 
the palaeography of vellum manuscripts) express the con- 
sensus of opinions of many experts and many ages. The 
whole subject is new ; fresh materials are coming to light 


year by year, and much of that which is already extant 
has not been published in such a form as to make it avail- 
able for students at a distance from the originals. Hence, 
although the wealth of the British Museum in papyri (and 
especially in literary papyri) gives a considerable advantage 
to a student whose work lies in that sphere, it is possible 
that the experts of Berlin and Vienna may sometimes have 
been led by their experience among the yet un-photographed 
documents in those collections to conclusions different from 
those which are expressed in the following pages. Still, 
the amount of accessible material is now so great, and 
spread over such wide periods of time, that inductions may 
be drawn from it with a fair amount of certainty; and 
with regard to the literary papyri, which form by far the 
most important branch of the subject, fortune has hitherto 
brought nearly every manuscript of the first importance 
to the British Museum ; so that it seems justifiable to try 
to state some general results and principles to which 
a study of these materials seems to lead us, in the hope 
that the development of this branch of palaeographical 
science may thereby be facilitated. 

In conclusion I have to thank the Delegates of the 
University Press for undertaking the publication of this 
volume, and especially Prof. Bywater for his supervision 
of it while passing through the press; and I wish also 
gratefully to acknowledge the assistance of Sir E. Maunde 
Thompson, K.C.B., Director of the British Museum, who 
has found time to read the proof-sheets and to make many 
valuable criticisms and suggestions. 

F. G. K. 

Octcber 22, 1898. 



I. The Range of the Subject ....... i 

II. Papyrus as Writing Material ...... 14 

III. Non-literary Papyri 34 

IV. Literary Papyri of the Ptolemaic Period .... 56 
V. Literary Papyri of the Roman Period .... 80 

VI. The Transition to Vellum 112 

Appendix I. Alphabets of Literary Papyri . . . 127 
Appendix II. Catalogue of Literary Papyri . . .129 
Appendix III. The Principal Publications of Non-literary 

Papyri 149 

Appendix IV. Abbreviations and Symbols . . . .154 




I. Letter, B.C. 242 i^Bodl. MS. Gr. class, c. 21 (P;) 
II. Petition, B.C. 161 (Brit. Mus. Pap. xliv) 

III. Loan, B.C. 105 ^rit. Mus. Pap. dclviii) 

IV. Official Letter, a.d. 15 (Brit. Mus. Pap. cclxxvi) 
V. Register of poll-tax, a.d. 72-3 (Brit. Mus. Pap 

CCLX) ....... 

VI. Receipt, a.d. 97 (Brit. Mus. Pap. cxliii) 
VII. Receipt, a.d. 166 (Brit. Mus. Pap. cccxxxii) 
VIII. Letter, circ. a.d. 350 (Brit. Mus. Pap. ccxxxvi) 
IX. Lease, 6tli century (Brit. Mus. Pap. cxiii. 3) 
X. Plato, Phaedo, 3rd century B.C. (Brit. Mus. Pap 


XL Dialectical Fragment, early 2nd century b.c 

(Louvre Pap. 2) .... . 
XII. Hyperides In Afhenogenem, and century b.c 
(Louvre Pap.) 

XIII. Bacchylides, ist century b.c. (Brit. Mus. Pap 

Dccxxxni) ...... 

XIV. Petition, circ. b.c 10 (Brit. Mus. Pap. cccliv) 
XV. Homer, Odyssey III, eafly ist century (Brit 

Mus. Pap. ccLXXi) ..... 
XVI. Hyperides In Demosthenem and Demosthenes, 
Olynthiac II (Rossall School Library) 
XVII. Lease, a.d. 88 (Brit. Mus. Pap. cxli) . 
XVIII. Herodas, ist-2nd centuiy (Brit. Mus. Pap. cxxxv) 
XIX. Homer, Eiad XIII, XIV, ist-2nd centui-y (Brit 

Mus. Pap. Dccxxxii) .... 
XX. Homer, Iliad II, 2nd century (Bodl. MS. Gr 

class, a. I. (P)) ..... 
Table of alphabets ....... 

to face 

page 37 

























,, lOI 

to follow page 128 




The science of palaeography in its application to Greek 
writing upon papyrus is a development of quite recent 
years. Sixty years ago. Greek palaeography practically 
began with the fifth century of our era, and dealt almost 
wholly with writing upon vellum and paper. Even the two 
great codices which now stand at the head of the list 
of Greek vellum MSS. were not at that time within the 
scope of the palaeographer ; for the Codex Vaticanus was 
jealously secluded in the Vatican Library, and the Codex 
Sinaiticus had not yet been discovered. It is true that some Previous 
examples of Greek writing upon papyrus had been known ^f ^^g*^'^ 
since the latter part of the eighteenth century : but these, subject. 
with the exception of the charred (and only very partially 
accessible) rolls from Herculaneum, did not contain literary 
works, and were merely isolated documents, or groups of 
documents, unconnected with one another or with the 
known palaeography of vellum MSS. Shortly before the 
middle of the present century, however, papyri began to be 
found in Egypt in somewhat more considerable numbers, 
and among them were several which contained works of 



literature. The discovery of a long roll, containing three 
of the lost orations of Hyperides, besides being nearly 
the first example of a literary MS. on papyrus in good 
preservation, was the first-fruits of the harvest of lost 
authors which Eg^^pt has yielded, and is still likely to 
yield, to the explorer. Some considerable portions of 
Homer were also added to the stock of available material 
at about the same time. But the progress of discovery 
was slow at first, and these earlier successes were not 
followed for several years by much that was important 
either in quantity or in (piality. It is only within the 
last twenty years that the stream has begun to flow 
with much fullness ; and only within the last seven years, 
or less, has it become possible to give anything like a 
continuous record of the character and development of 
Greek writing during the period when papyrus was the 
material mainly in use for its reception. 

The recent discoveries of papyri have, in fact, added 
a province of nearly a thousand years to the domain 
of the palaeographer, beginning about three hundred years 
before Christ, and coming down to a limit which is best 
fixed by the Arab conquest of Egypt in a. d. 640. It is 
only during the last three hundred years of this period 
that it overlaps the sphere occupied by vellum MSS. For 
six hundred years the papyri held the field alone, and only 
now are we beginning to be able to realize their character 
and understand the history of their development. That 
history has never yet been written with any fullness of 
detail ; even the outlines of it have hardly been laid down. 
When Gardthausen wrote his Griechisdie Paldograj^hie, in 
1879, his treatment of papyrus-palaeography was based 
upon a single group of documents belonging to the second 
century B.C., and two or three isolated and non-literary 
pieces scattered over the centuries between that date 
and the rise of vellum MSS. The great Hyperides MS. 
was the principal example of a literary hand accessible to 


]iim, and of that the age was quite doubtful. His treat- 
ment of the subject is, consequently, now out of date ; and 
no palaeograpliical treatise on the same scale has been 
written to take its place. The only work in which the 
new material has been dealt with is Sir E. Maunde 
Thompson's HainOiook of Greek and Latin Palaeography 
(1893), and that necessarily treats of it very briefly '. 
Moreover, even since the date of that book the mass 
of available material has been largely increased, and some 
serious gaps in our knowledge have been filled up. Under 
these circumstances, an attempt to set the material in 
order, and to formulate, even ;f it be but provisionally, 
the laws which regulated the development of Greek 
writing upon papyrus, seems to be justifiable. Though 
our knowledge is certainly destined to increase, and that 
soon, it may be useful to mark the limits which have 
hitherto been gained, and thereby to pave the way for 
future progress. 

The first discovery of Greek papyri in modern days was The oav- 
made at Herculaneum in 1752. Before that date the only co^veriesOf 
papyri known to the palaeographer were a few survivals piipyii- 
from mediaeval times. Letters, papal bulls, and municipal 
archives written upon papyrus in Latin existed in con- 
siderable numbers, chiefly in Italy; but in Greek little 
was extant at all, and that little was in extremely bad 
condition '^. The excavations at Herculaneum, however, 
brought to light a mass of charred papyrus rolls, to which 
must be assigned the honour of inaugurating the new era, 
though it was forty years before these began to be made 

' A still more brief survey of the subject is given by Prof. Blass in 
MuUer's Handbnch cJer Massischen Altertumswissenschaft, vol. i. (1891). The 
last edition of Wattenbach's Anleitung sur griechischm Paldographie (1895) 
attempts little more than a bibliography of the subject. 

2 Montfaucon, writing in 1708, says ' Graecum autem papyreum codicem 
nullum hactenus vidimus' {Palaeographia Graeca, p. 15), though he sub- 
sequently describes some much mutilated fragments in the libiary of 
St. Martin at Tours, and refers to the existence of a few othei's at Vienna, 

B 2 


available to scholars, the first volume of facsimile texts 
appearing at Naples in 1793. Even then their publication 
proceeded very slowly, especially during the troubled times 
of the next thirty years ; the Naples volumes were not 
very easily accessible to scholars in general ; and the 
charred condition of the rolls detracted greatly from their 
value. Meanwhile a far more important mine was opened 
in 1778 by the discovery in Egypt, probably in the province 
of the Fayyum \ of some forty or fifty rolls of papyrus. 
They were ofiered, by the natives who found them, to a 
dealer, who bought one out of curiosity, but refused the 
rest ; whereupon they were burnt by the natives for the 
sake of the smell which they gave forth. The one survivor 
was sent to Cardinal Stefano Borgia, and placed in his 
Museum at Velletri, where it was edited by N. Schow 
in 1788^. It contains nothing but a list of peasants 
employed in forced labour on the embankments regu- 
lating the Nile floods, and is of little importance save 
as the forerunner of much more valuable discoveries. 
For some time, however, discoveries came but slowly, and 
the publications of the next five-and-thirty years are few 
and unimportant. At last, about 1820, a new start was 
made. A large group of papyri was found (again, it is 
said, in an earthen pot) on the site of the Serapeum at 
Memphis; and these, which are now divided between the 
museums of Paris, London, Leyden, Rome, and Dresden, 
form the foundation of our knowledge of the palaeography 
of the second century B.C. Simultaneously papyri began 

' The find was reported to have been made at Gizeh, in a buried 
earthenware pot ; but since the surviving document relates to the 
local affairs of the village of Ptokmaidis Hormus, in the Fayj'um, it is 
much more likely to have been found in that neighbourhood. The 
statements of native discoverers as to the provenance of papyri are not 
valuable as evidence. 

* Charta papyracea Graece scripta Musei Borgiani Velitris (Kome, 1788), with 
specimen facsimiles. The document is of the end of the second century, 
being dated (though Schow overlooked this) in the latter part of the thirty- 
first year (of Commodus = A. d. 191). 


to appear from other sources, and the study of the materials 
thus provided soon led to the publication of several volumes 
of considerable note in the bibliography of papyri. In 
1 82 1 Mr. W. J. Bankes acquired a papyrus roll containing 
the last book of the Iliad ^ — the first literary papyrus to be 
discovered, with the exception of those from Herculaneum, 
In 1824 the publication of the Herculaneum papyri was 
resumed at Oxford. Two years later came the publication 
of the Turin papyri by Amadeo Peyron; while in 1839 
those of the British Museum were edited by Forshall, 
and in 1843 those of Leyden by Leemans. These three 
volumes contained nothing but non-literary documents ; 
but in the years just before and after the middle of the 
century several important literary papyri came to light. 
In 1847 Mr. Arden and Mr. Harris independently obtained 
portions of a great roll containing three speeches of 
Hyperides ^ — the first previously unknown classical author 
to be recovered on papyrus; in 1849 and 1850 Mr. Harris 
obtained two portions of a MS. of the seventeenth book 
of the Iliad, and another, in book form, of books ii-iv^; 
in 1855 M. Mariette secured for the Louvre a fragment 
of Alcman * ; in 1 856 Mr. Stobart acquired the Funeral 
Oration of Hyperides^. Some small portions of Homer 
were also acquired for the Louvre about the same time ^. 

With these discoveries the first age of papyrus-revelation Discove- 
may be said to have ended. A pause followed for twenty j.JJ^ "^ 
years, and then a new era was inaugurated, an era of twenty 
discoveries on a large scale, in which we may be said 
to be still living. In 1877 an enormous mass of papyri, 
mainly non-literary, was unearthed about the site of 

1 Commonly referred to as the Bankes Homer ; now Pap. cxiv in the 
British Museum. 

'^ Now Brit. Mus. Papp. cviii and cxv. 

3 Now Brit. Mus. Papp. cvii and cxxvi. The former (being the earlier 
to be made known) is the MS. generally known as tlie Harris Homer. 

* Louvre Pnp. 70. ^ Brit, Mus. Pap. xcviii. 

' Louvre Pap. 3, 3 ''*, 3 ''"'". 


Arsinoe in the Fayyum, of which the greater part went 
to the collection of the Archduke Rainer at Vienna, though 
Paris, London, Oxford, and Berlin also had a share. 

The papyri of this first Faj^yum find were almost entirely 
of the Bj^zantine period, and for the most part were very 
frao-mentary. Since that date, though papyri have also 
come from Eshmunen (Hermopolis), the neighbourhood of 
Thebes, and elsewhere, the mounds of the Fayyum have 
been the most prolific source of papyri, nearly all, however, 
up to the present, having been of a non-literary character. 
About 1892 another very large find was made in the same 
district, principally, it would appear, from the site of 
a village called Socnopaei Nesus ; and this time the lion's 
share fell to Berlin, although the British Museum secured 
a good representative selection, and others went to Vienna 
and Geneva. This second great find difiered from the first 
in containing far more perfect documents, and in belonging 
mainly to the Roman period, from the beginning of the 
first to the middle of the third century after Christ. 
Meanwhile the range of our knowledge had been extended 
backwards by Mr. Flinders Petrie's discovery (in 1889-90) 
of a number of mummy-cofiins, the cartonnages of which 
were composed of fragments of papyri written in the third 
century P.. c. Most of these were business documents ; but 
the literary specimens included tM^o very valuable items 
in the fragments of Plato's Fhaedo and the lost Aniiope 
of Euripides. About the same time the British Museum 
acquired a most remarkable series of literary papyri, in- 
cluding the lost W6r]vaioiv YloXirtia of Aristotle, the Mimes 
of Herodas, part of another oration of Hyperides, and 
a long medical treatise, to say nothing of MSS. of 
Homer, Demosthenes, and Isocrates. To these the Louvre 
added in 1892 the greater part of Hyperides' masterpiece, 
the Oration againd Athenogencs. In the same year a gap 
in the chronological sequence of dated papyri was filled 
by the acquisition on the part of the British Museum 


and the Geneva Library of a group of documents from 
the middle of the fourth century (the correspondence of 
a Roman ofRcer named Abinnaeus) ; while the papyri 
obtained by Messrs. Grenfell and Hog'arth in the years 
1894-1896 have provided ample material for the palaeo- 
graphy of the latter half of the second century b. c, and 
have even extended our knowledge for some distance into 
the previously blank first century b. c. The winter of 
1 896-7, however, threw into the shade all previous dis- 
coveries, with the possible exception of those of 1890-1. 
The British Museum acquired a papyrus containing con- 
siderable portions of the odes of Bacchylides; M. Nicole, 
of Geneva, secured some fragments, small but interesting, 
of a comedy by Menander; while Messrs. Grenfell and 
Hunt, excavating at Behnesa (Oxyrhynchus) on behalf of 
the Egypt Exploration Fund, unearthed masses of papyri 
which rival in quantity, and far surpass in quality, even 
the great finds of Arsinoe and Socnopaei Nesus. The 
thousands of papyri discovered on this occasion include, 
along with masses of fragments, large numbers of non- 
literary rolls in excellent condition ; and, more valuable 
but more tantalizing, scores of fragments of literary works, 
known and unknown. The full list of these discoveries 
has not yet been made public, but it includes the now 
famous ' Logia ' of our Lord, and fragments of St. Matthew, 
Sappho (probably), Thucydides, Sophocles, Plato, and other 
authors, besides the inevitable Homer. It may safely be 
predicted that the harvest of that season will not be 
surpassed for many a year to come. 

The general result, from the palaeographical point of Chrono- 
view, of this series of discoveries, and especially of those ([istribu- 
of the last six years, is that we now have an almost un- tionof tlie 
interrupted series of dated documents from B.C. 270 to 
A. D. 680. The third and second centuries B.C. are now 
so fully represented that there seems little room for 
serious error in dealing with MSS. of these periods. For 


the first century, especially the transition from Ptolemaic 
to Roman writing, material was until two years ago almost 
wholly wanting, and still is but scanty. From the beginning 
of the Christian era the supply begins to grow plentiful 
again, and from about the middle of the first century until 
the second quarter of the third century dated papyri exist 
in such numbers that hardly a year is without many of 
them. The last half of the third century is somewhat 
thinly represented ; but the age of Diocletian (a turning- 
point in the history of palaeography) is covered by some 
recent acquisitions of Mr. Grenfell's (now in the British 
Museum), and the middle of the fourth century is fully 
known from the papyri of Abinnaeus at London and 
Geneva, to which reference has been made above. A few 
documents dated about the year 400 have lately come to 
light, but the fifth century is now the darkest period in 
the palaeography of papyri. The sixth century and the 
first half of the seventh, on the other hand, have yielded 
a huge harvest; but the proportion of precisely dated 
documents is small, and the accurate apportionment of 
the various types of hands cannot be made with absolute 
certainty. With the Arab conquest of Egypt the practice 
of Greek writing on papyrus received its death-blow, and 
during the latter half of the seventh century it slowly 
flickered out. No dated document of the eighth century is 
in existence, except two tiny receipts in Arabic and Greek, 
though one long papyrus is assigned with apparently good 
reason to this period. 
Inferences The end of the papyrus period is, therefore, fixed and 
use of known ; but with the beginning it is different. Recognizing 
writing. ^}^at the material which we have from the early part of the 
third century b. c. is suflBciently plentiful to allow us to 
carry our inferences back for a generation or two, we may 
fairly say that we know how men wrote in the days of 
Aristotle and Menander; but we have not yet got back 
to Pindar and Aeschylus, much less to Homer or (if a less 


contentious name be preferred) Hesiod. There are, however, 
certain broad generalizations and inferences which have 
a bearing even upon these distant periods. In the first 
place it is clear that, at the point where our knowledge 
now begins, writing was a well established art, practised 
not only by literary men and professional scribes, but by 
soldiers, farmers, and working men and women of all sorts. 
It was used for the daily affairs of life, to all appearance, 
not less than it was in England a century ago. It could 
be written with ease and elegance ; it could also be written 
with a speed and fluency, which, if they sometimes offer 
serious difficulties to the decipherer, prove at least that the 
writer handled a ready pen. From this it is clear that it 
was no newly acquired art, but had already a long history 
behind it; how long, we have at present no sufficient 
evidence to say, but there is no sort of reason to be chary 
of generations. Further, it is obvious that the writing on 
papyrus bears no recognizable relation to contemporary 
inscriptions upon stone ; and therefore care is necessary in 
using epigraphic evidence to determine the style of writing 
in the preceding centuries. The characters used in inscrip- 
tions may resemble the formal writing (or printing) of the 
same age ; but they have no more bearing on the running 
hands in common use than modern tombstones have on the 
handwriting of to-day. Men may have carved formally and 
with difficulty upon stone ; it does' not follow that they 
did not write upon papyrus with ease and fluency. 

So far we have made no formal distinction between Distinc- 
literary and non-literary papyri ; yet the distinction is 


fundamental for the study of papyrus-palaeography. It literary 
is a distinction that runs through the whole period of ntera 


which we are speaking. Papyri which were meant to ^'litings. 
be books were written in quite different hands from the 
papyri which were meant to be documents, whether official 
or private. The difference is, in many cases, as marked 
as that between writing and print at the present day : 


nnd the development must be followed separately along 
each line. A parallel may be found in the distinction 
between the book -hands and the charter-hands of the 
Middle Ages. The charters of the twelfth century may 
bear some likeness, recognizable by the trained eye, to the 
books of the same period ; but from the thirteenth to 
the fifteenth centuries the relationship is practically indis- 
tinguishable, and a person who can date a charter of 
Edward III or Henry VII with certainty may be quite at 
sea with a chronicle or Bible of the same age. So with 
papyri, an acquaintance with the succession of non-literary 
hands only goes a little way towards enabling one to 
fix the date of a literary MS. And there is this further 
complication to be borne in mind, that whereas the 
charter-hand of the Middle Ages is the hand of a trained 
scribe just as much as the book-hand, the non-literary hand 
of the papyri includes the writings of private individuals, 
often very imperfectly acquainted with the use of the 
pen, as well as those of official clerks. We have, in fact, 
during the papyrus period what we have hardly at all 
in the vellum period of palaeography, the casual every-day 
writings of the common people ; and consequently the 
lines of classification which serve for vellum MSS. do not 
apply. For this reason the palaeography of papyri forms 
a branch apart, the principles of which must be stated 

Greek writing upon vellum can be classified in two 
broad and well-defined divisions, as uncial or minuscule, 
the former being the earlier, the latter the later style, — 
of course with some period of overlapping. Uncial writing- 
is never cursive, whereas minuscules are so commonly 
connected by ligatures that the terms minuscule and cursive 
are habitually used as synonymous, though they are not 
properly so. In papyri the circumstances are quite different. 
It is impossible to draw any distinction between uncial 
and minuscule ; and uncials, no less than minuscules, may 


be written cursively. An uncial hand without ligatures is 
not necessarily earlier than one which has tliem ; and, for 
non-literary purposes, hands ot" the most cursive character 
are found in the very earliest papyri yet discovered. The 
only classification which is of use in the study of papyri 
is that which has been stated above, into literary and 
non-literary hands ; and it will be necessary, in the present 
treatment of the subject, to deal with them separately. 

The material for the examination of these two branches 
is, however, of very different character. It is of the nature 
of a business document, such as a lease, a loan, or a receipt, 
that it should be accurately dated ; it is only rarely that 
a literary work will have any precise indication of the 
same kind. The long series of dated documents, spoken 
of in the preceding pages, consists almost exclusively of 
non-literary hands. Consequently our knowledge of non- Non-liter- 
literar}'- palaeography is far more exact than that which we graphy 
have of literary papyri. A non-literary document must 'tss "n- 
be written in a strange hand indeed if the doubts as to but bettor 
its period range over a hundred years, while those which ^j^^^j^" 
are written in anything like the hand of a trained clerk literaiy. 
can generally be placed approximately within the limits 
of a generation. But the dates assigned to literary MSS. 
liave fluctuated over several centuries, and cannot even 
now be fixed with absolute precision. Yet it is just 
of these MSS. that it is most important to know 
the age. The precise date of a petition from the fellah 
Sarapion to the magistrate Hierax is a matter which 
concerns a few specialists alone ; but the date of a MS. 
of Hyperides interests Greek scholars in general, and 
that of a copy of a Gospel, if one should be discovered, 
wovild be a matter of the gravest importance to theologians. 
The main object of the present essay is, consequently, to 
show how far science has progressed in this department ; 
to examine the whole series of extant literary papyri ; to 
show which can be dated accurately, and what are the 


probable dates of those which are still in doubt. Such an 
examination will have to deal with the subject practically 
de novo ; and it is hoped that this may be done without 
the smallest appearance of disrespect for the eminent 
scholars and palaeographers who have previously assigned 
dates to the extant MSS. It is a case in which an access 
of fresh material justifies a new comer in revising the 
work of his betters; and the purpose of the following 
pages is rather to show where new facts have come to 
light than to add to the number of conjectural dicta. 
Necessity In spite of the line which has l)een drawn between 
i'ngSi'. literary and non-literary papyri, it will not be possible 
to ignore tlie latter altogether in a study of the former. 
In the first place, the non-literary documents furnish us 
with certain broad criteria which are applicable to the 
literary documents. The main division, which will be 
made below, into Ptolemaic, Koman, and Byzantine periods, 
though more noticeable in the non-literary hands, is trace- 
able also with the literary; and certain forms of letters 
are common to both. But in addition to this the evidence 
on which literary papyri are dated not unf requently depends 
on our knowledge of non-literary palaeography. Titles, 
scholia, and corrections to a literary MS. are often written 
in non-literary hands, and so supply at least a terminus 
ante quem for the document in which they are found. 
In other cases, one side of the papyrus may contain literary 
writing, the other non-literary ; a knowledge of the manner 
in which papyrus rolls were made and written determines 
which of these writings is the earlier, and a knowledge 
of non-literary palaeography gives us a date to guide us 
in estimating the age of the more valuable writing. 

On all these grounds, then, it has seemed advisable not 
to limit this essay to the palaeography of literary papyri, 
but to include in it a sketch of non-literary writing as 
well. The latter will come first, because it is better 
known, and because the results of it are required for the 


examination of the less-known branch of the subject ; but, 
both because it is better known and because it is less 
important, it may be treated more briefly. The more 
important literary papyri, on the other hand, will need 
individual examination ; even small scraps of writing 
of this type have been noted as far as possible, since 
they will sometimes be found to throw valuable light 
upon the subject. But before proceeding to the palaeo- 
graphy, strictly so called, of papyri, it will be necessary 
to give some account of papyrus as a material for writing 
on, — the Biicliwet^en of the papyrus period, if a German 
word may be used where there is no exact equivalent in 
English. It will not be necessary to treat of this ex- 
haustively, since it would be useless to reproduce what 
is already accessible in the recognized handbooks of the 
subject ; but it will be useful to summarize what has been 
previously known, and to add the further information which 
is now available. Until recently, our knowledge of the 
subject has been almost entirely derived from the statements 
of ancient authors ; but we are now able to test and 
interpret these by an examination of the extant papyrus 
MSS. which the sands of Egypt have restored to us. 



Enrly use The use of papyrus in Egypt, the country of its production. 

iiiE^^^'t'^ goes back to an indefinite antiquity. The earliest extant 
specimen is a papyrus containing accounts of the reign 
of king Assa, whose date, according to a moderate estimate 
of Egyptian chronology, is about 3580-3536 b. c.^ ; while 
the earliest literary work which has come down to us (the 
Prisse Papyrus at Paris), although the copy which we have 
of it seems to have been written betw^een 2700 and 2500 B.C., 
is stated to have been composed (and therefore originally 
written down) in the same reign. In technical execution 
the papyrus rolls of the Egyptian kingdom leave nothing 
to be desired, and there is no reason to suppose that the 
manner in which the material was prepared differed at all 
from that of Greek and Roman days. We have, however, 
to pass over a period of some three thousand years before 
reaching a date at which we can be sure that papyrus was 
in use among the Greeks. The statement of Herodotus - 
that the lonians still applied the name of bLcpOepai to books 
made of papyrus, because they had formerly used skins 
as writing material, shows at least that in his time the 
Greeks of Asia used papyrus, and that its use was not 
quite a new thing. It is safe, therefore, to assert that 
it was the w^riting material of the Greeks at the beginning 
of the fifth century ; but how much further it can be 

' Petrie, History of Egijx>t, i. 8i. * Herod, v. 58. 


carried back must remain doubtful. Formerly it was 
held that the exclusiveness of tlie Egyptians in the mattt'i- 
of foreign trade made it inadmissible to argue from the 
practice of Egypt to that of Greece ; but the proofs of 
intercommunication have now increased so greatly that 
this argument has lost force, and we must be prepared for 
the possible appearance of evidence establishing the usi^ 
of papyrus by the Greeks at a much earlier period than 
has hitherto been held probable. 

With regard to the manner in which the writing material l'i<>para- 
was prepared from the plant, and the methods of writing niateriiil. 
upon it, not much has to be added to what has long been 
known on the subject and may be found in the recognized 
authorities, notably Birt, Gardthausen, and Thompson. 
The locvbs classicus on the preparation of the material is 
a well-known passage in the elder Pliny (N". H. xiii. 
] 1-13), the interpretation of which must be guided by the 
conckisions derivable from the papyrus documents actually 
in existence. The pith of the stem of the papyrus plant 
was cut into thin strips, the width of which was of course 
determined by the thickness of the stem, while their length 
varied considerably, as will be shown below. These strips 
(Lat. ph'dyrae) were laid side by side to form a sheet. 
Each sheet was composed of two layers, in the one of which 
the strips ran horizontally, while in the other they were 
perpendicular. The layers were attached to one another 
by glue, moistened with water, — preferably, it would appear, 
the turbid water of the Nile, which was supposed to add 
strength to the glue \ The sheets thus made were pressed, 
dried in the sun, and polished, so as to remove unevenness 
in the surface ; and they were then fit for use. 

' It is almost certain that this is the true sense of Pliny's 'turbidus 
liquor vim glutinis praebet,' as held by Birt and Thomp.son. Apart from 
the inherent improbability of the Nile water being able by itself to act as 
glue, it may be added that traces of glue are visible in extant papyri. 
The amount used was, no doubt, very small, so as to avoid aflecting the 
surface of the papyrus. 


Size of the The size of the sheets (KoXki'uxaTa) varied according to the 
quality of the papyrus, only one rule being constant, that 
the height is greater than the breadth, when the sheet 
is held in the way in which it is meant to be used. Pliny 
(/. c.) gives the dimensions of the different qualities known 
in his time, and his figures were formerly supposed to 
apply to the height of the papyrus. This view is, however, 
invalidated by the testimony of the extant papyri, and 
Birt is no doubt right in referring Pliny's measurements to 
the width. The largest size named by Pliny (which was also 
the most valuable) is 13 digiti, or about 9I inches; but 
papyri are extant which are as tall as 15I inches, and 
from JO to 12 is quite common in documents which make 
no pretence of special handsomeness or excellence. Further, 
we do not find that poor people necessarily use papyrus 
of very small height ; while it is true that narrow sheets 
are often used for comparatively unimportant purposes. 
Thus, to quote cases in which only a single sheet is required 
for a document, the two most elaborately written petitions 
in the British Museum (Papp. CCCLiv and clxxyii), which 
are addressed to the highest official in Egypt, measure 
respectively 8 J and 6} inches in width; while receipts, 
records of loans, and the like, are very commonly written 
on papyrus not more than three inches wide. An examina- 
tion, moreover, of such long rolls as are at present known 
tends to confirm Birt's view that Pliny's dimensions refer 
to the width of the KokXr}ixaTa. The finest literary papyrus 
in existence, the British Museum Odyssey (Pap. cclxxi) 
has KoWrijxaTa of just over 9 inches in width ; while in the 
Bacchylides papyrus, which is likewise a handsome roll, 
they vary between 8 and 9 inches. In the Herodas MS., 
which is small in height and unostentatiously written, they 
are only 6 inches in width. The papyrus of Hyperides In 
PliilijJpidein and Demosthenes' Third Epistle, which is only 
gl inches in height, has KoWijixara 72 inches wide; while 
in a tax-register (Brit. Mus. Pap. cclxviii), which reaches 


the extraordinary height of 15^ inches, they are only 5 
inches wide. The papyrus of the 'AOi^vaCodv FIoAtreta, which 
was originally intended merely for a farm-bailiff's accounts, 
has KoXKi'ifxaTa of 5 to 55 inches in width ; and this is a very 
common size for non-literary documents. 

For non-literary documents, siich as letters, receipts, 
deeds of sale or lease, contracts, petitions, and the like, single 
sheets of papyrus could often be used; but for literary 
purposes a number of sheets were united to form a roll. 
According to the ordinary interpretation of Pliny's words, 
not more than twenty sheets went to a roll ; but this 
statement is not confirmed by the extant papyri. Twenty Forma- 
sheets of even the widest kind (gl inches each, as stated in j.o1"g°f 
the last paragraph) would only give a roll of 16 feet; and papyrus, 
this length was certainly often exceeded. Egyptian papyri 
sometimes run to enormous lengths, in one case to as much 
as 144 feet; but these need not be taken into account. 
They were for show rather than for use, — editions de luxe 
which the owner proposed to take with him to the next 
world, where he might have strength to grapple with them, 
but which he certainly did not want to read on earth. 
But even Greek papyri, though they do not approach these 
dimensions, often exceed the limit which the text of Pliny 
appears to assign to them. From 20 to 30 feet may be Their 
taken as a full size, the higher limit being rarely, if ever, ^"^^"S^^^- 
exceeded \ The largest papyrus of Hyperides (Brit. Mus. 
Papp. cviii, cxv), containing the orations against Demo- 
sthenes and in defence of Lycophron and Euxenippus, 
must have measured, when complete, about 28 feet; a MS. 
of the last two books of the Iliad (Brit. Mus. Pap. cxxviii), 
about 25 feet; and that of the Mimes of Herodas (Brit. 
Mus. Pap. cxxxv) perhaps about the same. On the other 

' The papyrus containing the Eevenue Laws of Ptolemy Philadelpliua, 
obtained by Mr. Petrie and now in tlie Bodleian, measures 42 feet ; but 
this is composed of several distinct documents attached to one another, 
and, moreover, is not a literary work. 



hand, a papyrus containing the De Pace of Isocrates (Brit. 
Mus. Pap. cxxxit) measures 14 feet; while that of Hyperides 
In Athenogene'ni cannot have measured, when complete, more 
than about 7 feet. The true interpretation of Pliny's state- 
ment, no doubt, is that in practice the sellers of papyrus 
kept and sold it in lengths (Pliny's scajn) consisting of twenty 
sheets. Egyptian rolls have been observed in which the 
number 20 is marked at the end of each twentieth Ko'AArj/xa, 
and this no doubt indicates the end of each length of papj^rus 
as purchased by the author from the stationer ^. But the 
author was no more limited by this fact than the modern 
writer is limited by the fact that he purchases his foolscap 
by the quire or the packet. He could join one lengtli of 
papyrus on to another, and when he had finished his work he 
could cut off whatever papyrus w^as left blank. It is not in 
the least probable a priori that there was any hard and fast 
rule fixing the length of a papyrus book, nor do the facts 
disclosed by the extant MSS. authorize such a supposition. 
Their The height of a papyrus varies considerably, but the 

average may be taken to be from 9 to 1 1 inches. The 
tallest at present known is a tax-register in the British 
Museum (Pap. ccLxviii), which measures i^h inches. Three 
census rolls (Papp. cclvii-cclix) measure 13^ inches; and 
about : i inches is quite a common height. Literary 
MSS. are generally rather smaller. Of those mentioned 
above, the principal Hyperides MS. measures 12 inches, 
the Isocrates 1 1 inches, the Homer 9 1 inches ; while 
the Herodas (which must be regarded as a kind of 
pocket volume, such as volumes of poetry often are in 
modern times) is only 5 inches in height. The Louvre 
Hyperides measures 9 inches, the British Museum MS. 
of the same author's Oration againd Pldlippides <)\ 
inches ; while the two most handsome literary papyri 
now extant, the British Museum Odyssey and Bacchylides, 

' See Borehardfc, Zeitschrift fur (igyptische Sprache xxvii. 120, Wilcken, 
Hermes, xxviii, 167. 



measure respectively 13 and gf inches. To give the oldest 
extant examples, the Petrie Phaedo measures 8^ inches in 
height, and the Antiope 85 inches. 

The writing was normally on that side of the papyrus Fiedo anrT 
on which the fibres lay horizontally (technically known ^^'^^°' 
as the recto) ; and this is a rule of much importance ; for 
when, as is frequently the case, a papyrus has been used 
on both sides, it is often only by this rule that it can be 
determined which writing is the earlier. It is therefore 
necessary to state the law somewhat precisely, following 
Wilcken, to whom the first formulation of it is due ^. It 
is obvious on reflection that, by holding the papyrus 
differently, the fibres on either side can be made to lie 
horizontally, only in one case they are parallel to the 
height of the roll, and in the other to its length ; but it 
does not follow that either side can be made the true redo 
at will. The true redo of the original sheets of papyrus 
out of which a roll is made is that side on which the 
shorter fibres are (probably because greater perfection and 
evenness could be secured with short fibres than with long 
ones) ; and the right way of holding such a sheet is to make 
these fibres lie horizontally (since thus the least obstacle 
is offered to the pen). Hence, when several such sheets 
are joined together, side by side, into a roll, all the fibres 
on the side which is primarily intended for writing will lie 
horizontally ; and, conversely, the side of a roll on which 
the fibres run in the direction of the length of the roll is 
that which is primarily intended for writing. If writing 
is found running at right angles to the fibres, one of two 
things must have happened : either the scribe has written 
on the verso of the papyrus, or he is holding the redo in an 
unusual way. Examples of both are found. The verso 
is used when the redo already has writing upon it, or, 
occasionally, in the Ptolemaic period, without obvious cause ; 

^ 'Kecto oder Verso,' in Htrmes, xxii. (1887). 

c a 


iieAer, so far as present experience goes, in Roman times, 

unless the recto has been previously used. Writing on the 

redo, but at right angles to the fibres, is found in a few 

Ptolemaic documents and in many of the Byzantine period. 

In the former cases the roll has been made up in an unusual 

manner, the sheets being joined together top and bottom, 

instead of side by side ; so that the writing, though it is 

across the fibres, still runs parallel to the length of the roll. 

In the Byzantine documents, on the other hand, the roll is 

made in the ordinary way, and the writing runs parallel to 

the height of the roll, so that in reading the roll has to be 

unfolded from the top downwards, instead of sideways. 

This method is not applied, however, to works of literature. 

The general rule, — invariable in the case of literary 

works, apparently invariable in non-literary works of the 

Roman period, and largely predominating in non-literary 

works of the Ptolemaic period, — is that the first writing 

on the papyrus is parallel to the fibres. The exceptions 

can easil}^ be recognized after a little experience, — in the 

case of large papyri, by looking for the traces of the 

original sheets, and in the case of small scraps by noticing 

the greater smoothness of the true redo. It is rare to 

find a work continued from the redo to the verao. A lono- 

book of magical formulae (Brit. Mus. Pap. cxxi) and an 

account book (Pap. cclxvi) are examples of this practice ; 

but ordinarily if there is writing on the verso it is quite 

independent of that on the redo. It may safely be assumed 

that no MS. of a literary work intended for sale 

was ever written on the verso ; though copies for private 

use might be, as in the case of the \\di]vai(iiv rToAtreia of 

Aristotle. In a long medical papyrus in the British 

Museum (Pap. cxxxvii), and in the Revenue Papyrus of 

Philadelphus, notes or additions to the main work are 

written on the verso ; but neither of these is a literary 

work intended for sale. 

Width of The normal method of arranging the writino: on a 
columns. & o to 


papyrus roll was in columns, the lines of which run 
parallel to the length of the roll, as above described. 
For literary MSS. this method is invariable ; the Byzantine 
documents mentioned above, in which the writing ran in 
one large column down the whole papyrus, in lines at 
right angles to its length, are of a non-literary character, 
generally wills or leases. The width of these colunnis 
(o-eAiSes) varied, but for literary MSS. intended for sale 
the length of a hexameter line may be taken as determining 
the extreme width. This, in a hand of good size, implies 
a width of about 5 inches, besides the margins, which 
might be as much as i| inches between the columns and 
2 or 3 inches at the top and bottom ^ In prose works, 
so far as our present knowledge goes, the width of the 
columns is generally much less. The widest are found 
in the Louvre papyrus of Hyperides, which measure about 
3I inches; the narrowest in one of the British Museum 
papyri of the same author (the In FhUippidem), which 
are barely half that width (if inches). The large 
Hyperides papyrus has columns 2 inches wide, while 
those of the British Museum Isocrates measure a little 
less than 3 inches. The only literary papyrus in which 
these dimensions are exceeded is that of Aristotle's 
'M-^vaiMv Ylokirda, which has one column measuring as 
much as II inches wide, while several others are 5 or 
6 inches; but this does not constitute a real exception, 
since the MS. is not written in a literary hand, nor 
intended for publication 2, In non-literary papyri much 

^ These are the measurements in the British Museum paj^yrus fragment 
of the Odyfisnj (Pap. ccLXXi -, which may be i-egarded as the handsomest 
literary papyrus at present extant. In anotlier well-written Homer MS. 
(Brit. Mus. Pap. cxxviii) the figures are slightly less. 

2 It has sometimes been supposed that the ceXtSfs correspond with the 
KoWrj/iara, that is, that the writing was not allowed to cross the junctions 
between the sheets of which the papyrus roll was composed ; but this 
is not borne out by the facts. In the best-written MSS. (such as the 
British Museum Odyssey and Bacchylides), no less than in the worst, 
the writing fi'equently crosses the junctions. 


o-reater widths are sometimes found, their dimensions 
being in fact determined by their contents. In census- 
lists and some kinds of accounts the scribe preferred to 
get each entry into a single line; hence in a document 
of the former class (Brit. Mus. Pap. cclx) some of the 
columns are lo inches in width, while in a tax-register 
(Pap. cxix) they are as much as 12 and 13 inches. 
Ptolemaic scribes had a fondness for writing such things 
as wills, leases, loans, and the like, in one or two very 
broad columns. To give two instances only, Brit. Mus. 
Pap. DCLXXV is written in a single column measuring 
15 inches in width and only 6 inches in height; while 
Pap. Dcxxiii is in two columns of 13 inches, besides 
a short al^stract of the contents in a separate column. 
These, however, are only examples of a temporary fashion, 
which must have been found inconvenient in practical use, 
and could never have been adopted for any composition 
which was likely to be often read. 
Titles. The mutilation of nearly every literary papyrus which 
has come down to us renders it difficult to lay down any 
very certain rule as to the methods commonly employed 
by the ancients to indicate the contents of a roll. It was 
certainly not unusual to inscribe the title of a work at 
the end of it, as is the case with the largest Hyperides 
MS., the British Museum Isocrates, and several of the 
papyri of Homer; but it was not invariably done, for it 
is not found in the MS. of the In Philippidem., the end 
of which is perfect, nor yet in the Aristotle papyrus. It 
is also obvious that it would have been extremely incon- 
venient to have to unroll the whole of a volume in order 
to see what its contents were. It was usual to leave a 
blank column at the beginning of a roll, as is found in the 
papyri of Aristotle and Herodas and in the Harris Homer ; 
but in none of these cases is the title written on it ; and 
in the great Hyperides MS., where the title was so written, 
it is by a different and apparently later hand. It seems 


certain, therefore, that the ordinary way of indicating the 
title of a work was by the aikXv'jios, or little strip of papyrus 
or vellum attached to and projecting from the roll; and 
these, though known from the references in ancient 
authors to have existed ^ have in no case come down to us. 

In all palaeographical works it is stated that the roll, Use of 
when completed, was rolled on a stick (ofi(/)aAoj), ornamented ^''''^®'"^- 
at the ends with projecting knobs or tips (K^para) ; and 
the statements of Latin writers ^ leave no doubt that this 
was the habitual practice in the case of their works. 
The actual papyri which have come to light of recent 
years make it necessary to modify this proposition. In 
no case (except in dummy rolls manufactured for sale to 
tourists) has a wooden roller been found ; many of the 
Herculaneum papyri had a central core of papyrus ; some 
burnt rolls brought from Egypt a few years ago had, in 
some instances, a reed or quill in the middle ; but as a rule 
there is no trace of any roller at all. This fact is perfectly 
intelligible in itself, and may quite well be reconciled with 
the statements of the Latin poets. Papyrus was not 
originally the brittle material which, from its appearance 
after the lapse of a score or so of centuries, one is apt to 
imagine, and could quite easily be rolled upon itself; and 
for this purpose, as well as to resist tearing, the ends of 
the roll are often stiffened by an extra thickness of papyrus, 
about an inch in breadth ^. In the case of common copies 
this was, no doubt, the regular practice ; while the 
handsomer books would be provided with wooden rollers 
and all the other appurtenances of style and luxury. The 
distinction would be very much the same as that between 
cloth-bound and paper-covered copies at the present day. 
The Latin poets are speaking of the dainty copies of their 
works which were to be seen in the bookshops and salons 

^ Thompson, Chxek and Latin Palaeography, p. 57. ^ lb. p. 56. 

' A good instance may be found in the Harris Homer (Brit. Mus. 
Pap. cvii). 




of the capital ; wliile the papyri which have come down 
to us are generally from the houses and tombs of obscure 
provincials in Upper and Middle Egypt. 

On the further details of book production in the papyrus 
period there is nothing new to be recorded. The allusions 
in the Latin poets provide us with all we know as to 
the ^atro'A?;?, or wrapper, with which the roll might be 
covered, the cedar-oil by which it might be protected 
against insects, the chest (resembling a bucket, to judge 
from extant representations) in which it was kept. These 
details have already been gathered together in the recognized 
handbooks of palaeography, and it seems useless to repeat 
them here. The discoveries of actual papyrus rolls have 
added nothing to our knowledge on such points. 

Piipyrus The description of a papyrus book, which has been 
given in the preceding pages, applies to nearly the whole 
of the papyrus period, as at present known. But towards 
the end of the period the codex, or modern book form, 
is found coming into existence side by side with the 
traditional roll form. The origin of the codex is, no 
doubt, to be found in the sets of wax tablets which were 
in use for note-books at least as far back as the first 
century B.C., and probably much earlier. These tablets, 
consisting of wax laid upon wood and surrounded by 
raised wooden rims, were bound together by strings or 
leather bands passing through holes bored in the rim 
of one of the longer sides of each tablet, so as to form 
something in the shape of a modern book. It was not 
until several centuries later, however, that this shape 
was adopted for literary compositions. As Sir E. Thompson 
has pointed out ^ the cause of the final victory of this 
form was that it was possible to include so much more 
matter in a codex than in a roll ; and the requirements 
of the churchmen and the lawyers agreed in giving it 

' Greek and Latin Palaeography, p. 61, 


the preference. A single Gospel was as much as a papyrus 
roll could contain, while a vellum codex could include 
the entire BiUe; and the great legal collections of the 
reign of Justinian would have needed a whole library 
of rolls. For a short time the experiment was tried of 
making papyrus codices, the papyrus being cut up into 
leaves and fastened together by strings or leather bands. 
It does not appear, however, that the experiment gave 
satisfaction, for very few instances of such papyrus l)Ooks 
have come down to us. A MS. of the second, third, 
and fourth books of the Iliad, formerly assigned to the 
fifth century, but, for reasons which will be given sub- 
sequently, more probably belonging to the third (Brit, 
Mus. Pap. cxxvi); a page of Menander at Geneva, 
probably of somewhat later date ; the Berlin fragments 
of the 'AOrivaLMv rioAtreia, probably of the fourth century ; 
a copy of the prophet Zechariah, of the sixth or seventh 
century, exhibited at the London Oriental Congress in 1892, 
and last seen in a dealer's shop in Vienna; some leaves 
of a Psalter of the same date, in the British Museum (Pap. 
xxxYi) ; small portions of two Hesiod MSS. at Paris and 
Vienna, of the fourth or fifth century ; a magical papyrus 
in the British Museum (Pap. xlvi), and another in the 
Bibliotheque Nationale (MS. Suppl. grec. 574), both of the 
fourth century ; these are the principal examples of Greek 
papyrus codices at present known to exist. There are 
Coptic volumes of this kind of much greater size, resembling 
the large vellum quartos and folios, but there is no sign 
that these were ever adopted for Greek literature. The 
rise of the codex was accompanied by the rise of vellum, 
and the papyrus period, so far as Greek literary works 
are concerned, was then coming to an end. 

Before beginning the history of Greek writing on Accents, 
papyrus, and of the evolution of the written characters jngs^ and 
individually, it will be as well to dispose of some subsidiary ''^''^^- 


uuitters, such as punctuation, accentuation, and the use 
of breathings. In all these respects papyrus MSS. are 
in a very elementary stage. None has a full equipment 
of stops, accents, and breathings; many have none at all. 
Further, the use of them does not follow regular laws 
of development ; or rather, the materials now available 
do not allow us to ascertain any law. The probability 
is that, the higher the quality of a MS., the fuller is 
its equipment in these subsidiary guides to intelligence. 
Non-literary MSS. very rarely have any of them. 
Accents are not found in them at all ; breathings in 
extreme rarity ; and only a few have some show of 
division of sentences. In the lower classes of literary 
MSS. an occasional accent is found, probably when there 
was some likelihood of a mistake as to the meaning of 
a word ; breathings are equally rare ; and only important 
breaks in the sense are indicated by punctuation-marks 
or blank spaces. Of the highest class of MSS., those 
which were intended for sale or for preservation in large 
libraries, there are very few extant specimens ; but 
there is some indication that accents, breathings, and 
punctuation-marks were more freely used in them, though 
without any approach to the completeness of later usage. 
Separation The most elementary form of assistance to the reader 
consists of the separation of words from one another. 
Where this exists, it is not usually very difficult to deter- 
mine for oneself the pauses in the sense. But, perhaps 
because it is so elementary, it is the last form of assistance 
to be given in Greek MSS. There is sometimes an 
approximation to it in non-literary papyri, where, the 
text being written cursively, the writer not unnaturally 
lifts his pen of tener at the end of a word than elsewhere ; 
l)ut this is so irregular and incomplete as to furnish very 
little help. In literary papyri the separation of words 
is almost wholl}^ wanting ; perhaps the only example of 
it is in a short grammatical treatise, bearing the name 


of Tiyphon, written not earlier than the fourth century 
on some blank pages in a MS. of Homer in the 
British Museum (Pap. cxxvi). In other MSS. the nearest 
approach to such a practice is the use of a dot, above 
the line, to indicate the true word-division in cases where 
the reader might easily make a mistake at first sight. 
Thus Brit. Mus. Pap. cxxvi gives, in Homer, II. iii. 379, 
a\\r • iTTopovae ^ : and a similar system is found in the Mar- 
seilles papyrus of Isocrates. It is however rare, the dot, if 
used at all, being generally required to separate sentences 
rather than words. A comma below the line is also found 
occasionally for this purpose (e.g. Bacch. xvii. 102). 

Punctuation in the ordinary sense of the word, or the Punctua 
indication of a break in the sense, is more common, though ^^"' 
still only sporadic. The earliest system would seem to be 
that of leaving a slight space in the text, and placing a short 
horizontal stroke {TTapdypa(pos, more rarely '7:apaypa(f>ri) below 
the beginning of the line in which the break occurs ^. This 
use of the Trapaypac^o? is mentioned by Aristotle ^, and is 
found in some of the earliest extant papyri. Thus in the 
fragment of the Antmpe among the Petrie papyri (third 
century B. c.) it is used to indicate the end of each actor's 
speech ; and similarly, along with the double dots mentioned 
below, in the Petrie Phaedo fragment. It is found also 
in the Louvre Hyperides (second century b. c), the British 
Museum MSS. of the same author (first century B. c. and 
first century after Christ), the Herodas (first or second 
century), and in several other MSS. In the Bacchylides 
MS. (first century b. c.) it marks the end of each strophe, 
antistrophe, and epode. Mistakes are sometimes made 
by transcribers in placing the TTapdypa<po'i, but its proper 
place is below the line in which the pause occurs. It 
marks the end, not the beginning, of a sentence. 

^ Classical Texts from Papyri in the British Museum, p. 82. 

^ Spaces in the text, without paraj/cap/ii, ai-e found in some literary papyri 
(e. g. the Herodas MS. sometimes), and not unfrequently in non-litei-ary 
papyri, especially those of a legal nature. ^ Rhet. 3. 8. 1409 a. 20. 


The use of dots for the same purpose is equally old. In the 
Petrie Fhaedo, which is of the same age as the Antioi^e, and 
in the Vienna papyrus containing the ' Curse of Artemisia,' 
which may be older, a double dot resembling a colon is 
used to separate sentences. This occurs also in the Erotic 
Fragment found and published by Mr. Grenfell (now Pap. 
DCV in the British Museum), which is of the second century 
B. c. ; but it is not common. The single dot, generally placed 
well above the line, is common. It is used freely, and 
almost regularly, in the Bacchylides papyrus, occasionally 
in the British Museum Odyssey MS. (Pap. cjlxxi); and 
it has been added by later hands to the three earliest 
papyri of the Iliad in the British Museum (Papp. cvii, cxiv, 
cxxviri). The use of the dot is known to have been 
systematized by the Alexandrian critics (traditionallj^ by 
Aristophanes of Byzantium), and different values were 
assigned to it according as it stood above the line (a full 
stop), in the middle of the line (a comma), or on the line 
(a semicolon) ; but this system cannot be traced in the 
extant papyri, where the dot is generally above the line 
(practically never on it), and is used to indicate minor 
pauses, such as a semicolon or even a comma, quite as much 
as for a full stop. 
Aocentua- Accentuation is rarer than punctuation in Greek papyri, 
and quite as fluctuating in its appearance. It is not found 
at all in non-literary documents, and in literary MSS. its 
use is sporadic. It does not appear in the Petrie papyri 
of the tliird century b. c, nor in the Louvre Hj-perides of 
the following century. The earliest example of the use 
of accents is in the Bacchylides MS., where they are also 
more plentiful than in any other papyrus ; and it is worth 
noticing that this is likewise one of the most carefully 
written papyri in existence, and is therefore probably 
something more than a copy for private use. Accents are 
also somewhat freely used in the Alcman fragment in the 
Louvre, which is probably of the latter part of the first 



century B. c. ; and more scantily in the Britisli Museum 
Odystey papyrus (early first century) and two fragments of 
the Iliad in the Louvre (first and second centuries). The 
Harris and Bankes papyri of the Iliad, which will be found 
placed below in the first and second centuries, though the 
former has hitherto been held to be older, have many 
accents, but not by the first hand, so that their date is 
uncertain ; and the same is true of the Britisli Museum 
papyrus of the last two books of the Iliad (Pap. cxxviii). 
In the much later MS. of Iliad ii-iv (Pap. oxxvi), which 
is probably of the third century, the accents are by the 
first hand. None of the Hyperides MSS. in the British 
Museum has accents, and the Herodas only a few isolated 

It is thus clear that accents are not to be looked for in 
papyri with any confidence, and are never used to the full 
extent that has since become customary. Even fairly well 
written MSS., such as the Petrie Phaedo and Ardiope, the 
Hyperides papyri, and the British Museum Iliads (with 
the exception of the latest), have none by the first hand. 
It is perhaps significant that the only two texts (earlier 
than the third century) in which they are at all largely 
used are both of them lyric poets. It may well be that in 
Bacchylides and Alcman the scribes felt that the reader 
required more assistance than in Homer or Hyperides. In 
these MSS. accents are especially applied to the longer 
words, and particularly to compounds, which are somewhat 
misleading to the eye. Prepositions, articles, pronouns, 
and adverbs very rarely have them, unless there is some- 
thing abnormal about them, as when a preposition follows 
its case {e.g. Kparos vTrep in Bacch. xviii. 51). In the case 
of diphthongs, the acute accent is generally on the first 
letter, and the circumflex over both, contrary to modern 
usage. In the Bacchylides the accent is never placed on 
the final syllable in oxytone words, but the preceding 
syllables have the grave accent; e.g. TTayKpanjs, Kparos. 


The explanation of this is to be found in the original 
theory of accentuation, according to which every syllable 
has an accent, but only one in each word is acute, the rest 
being grave ; e. g. avdpcairos, KaphCa, should be written 
av0po)TTo^, Kapbia. In practice the grave accents were 
omitted ; but why they were revived, to the exclusion of the 
acute, in oxytone words, is not clear. Traces of the same 
practice are found in the Harris and Bankes Homers ; 
€. g. eXcov, (fypeaw, and (in a proparoxytone word) enea-a-^vovTo^. 
The general principle governing the use of accentuation 
in the papyrus period would seem to be that accents were 
only inserted if the scribe felt they were wanted as an aid 
to reading, and (so far as yet appears) solely in texts of the 
poets. Just at the end of the period they came into more 
general use, and were sometimes sujDplied to previously 
existing MSS. ; but at this stage the supersession of 
papyrus by vellum came about, and accents once more 
disappeared from Greek texts for some hundreds of years. 

Breath- Much the same may be said of breathings, which are 
usually found in the same MSS. as accents. No papyrus 
is early enough to show the letter H in its original use as 
an aspirate; but the two halves of this letter, |- and -\. 
indicating the rough and smooth breathings respectively, 
are found in the Bacchylides MS. (though not uniformly), 
and in a few instances in the British Museum Odyssey 
papyrus. The more usual forms, however, both in these 
MSS. and elsewhere, are ■- and -" , or r and "i . The rounded 
breathing is not found in papyri, though the inverted 
comma (') is used as a mark of elision in the Bacchylides 
and other MSS. As with the accents, breathings are only 
used intermittently, when the scribe thought them neces- 
sary in order to avoid confusion or mistake. 

Other Marks of diaeresis (••) are often used over t and v, 


especially at the beginnings of words. They are found in 
non-literary as well as literary papyri. 

1 Catalogue of Ancient MSS. (Greek) in the British Museum, pp. i, 6. 


Dots are sometimes placed over letters, to indicate that 
they are cancelled. This is especially found in the large 
Hyperides MS. and the Herodas ; elsewhere it is more usual 
to draw the pen through the cancelled letters. Corrections 
are normally written between the lines, above the words 
for which they are to be substituted ; occasionally (e. fj. in 
the Aristotle papyrus) they are inclosed between two dots, 
but this is unessential. Omitted lines are supplied in the 
upper and lower margins, with a mark at the place where 
they are to be inserted {e.g. Bacchylides and Herodas 
MSS.). The margins (lateral as well as upper and lower) 
are also the place for scholia {e.g. Odyssey papyrus, Brit. 
Mus. Pap. cxxviii, etc.). 

Other marks, such as 7 or = to fill up blank spaces at 
the end of a line (Brit. Mus. Hyperides MSS.), a hyphen 
{^) under compounds to show that they are single words 
(Bacchylides MS.), and the like, are apparently due to the 
fancies of the individual scribes, and explain themselves. 
Two of the critical marks used by Aristarchus to indicate 
spurious or repeated lines in Homer (the lirX)], >-, and 
asterisk) are found in Brit. Mus. Pap. cxxviii {11. xxiii, 
xxiv), the Oxford papyrus of II. ii, and perhaps in Brit. 
Mus. Pap. ccLXXi {Od. iii) ; but they are not fully inserted 
in an}^ of these MSS. 

A brief mention should be made of the rules for the Division 
division of words in Greek papyri, because the point is at end 
often of importance in the restoration of mutilated texts. ^^ ^^"'^• 
If a word has to be divided at the end of a line, the rule 
is that the division should be made after a vowel, except 
in the case of doubled consonants, where it is made after 
the first consonant, or where the first of two or more 
consonants is a liquid or nasal, when it is divided from the 
others. Thus, in the course of a few lines of Hyperides 
Pro Euxenippo the following examples occur: 'ik€\yov, 
8^|/xw, hi\Ka(TTripiov, hiKa\(TTai, aAJAot, e^oi'lres. In the case 


of words compounded with a preposition, the division 
is most commonly made after the preposition; but not 
nnfrequently the normal tendency to make the break at 
a vowel prevails. Thus we have d(T\ayy^XXo[xivu)v, 7rpoo-|- 
?>ez', but also d\<jayyekias , v\T:^Xa[x^ave^. The same tendency 
is seen at its strongest in such divisions as rav\T\ oI\k, 
Ka\6\ which are far from uncommon. Breaches of these 
rules are practically unknown in literary papyri ^ The 
only point in which variation is admitted is in combina- 
tions of a with another consonant, some scribes making the 
division before the a, and some after it. Thus in some 
MSS. we find StKco-jrat, and in others hiKa^TTai Occasionally 
the same scribe will fluctuate in his practice in this respect. 
With regard to non-literary documents it is dangerous to 
assert an universal negative, on account of their great 
number ; but a special search through a considerable body 
of them has failed to find a single example of a division 
contrar}^ to the principles above stated, and it is at least 
clear that the rule is so generally observed that any breach 
of it must be regarded as quite exceptional. 

Abbrevia- Abbreviations are not found in well-written literary 
*^'^"''" papyri, with the exception of two Psalters (late third 
and seventh centuries), in which the common compendia 
for Kvpioi, Oeos, k.t.X., occur. In literary texts written in 
cursive hands, however, they are found somewhat largely 
in use. The chief example is the papyrus of the ' M-^valoiv 
rioAiTeia, where two of the writers whose hands occur in 
the MS. use twenty-five abbreviations for common words 
such as articles, prepositions, Ka'i, ovv, elvat, k.t.X. The large 
medical papyrus in the British Museum (Pap. cxxxvii) 
uses nearly as many, some, but not all, being identical. 
One of the Hei'culaneum papyri of Philodemus (Pap. 157- 

1 The contrary is stated by Wattenbach {Griech. Pa'uogr. 3rd ed., pp. 15, 
118 , who characterizes the division of words in Egyptian papyri as 'ganz 
regellos ' ; but this is quite a misconception. 



152) has a few abbreviations ', and there are also a few 
in the scholia to the Aleman fragment, and in a collection 
of rhetorical exercises in the British Museum (Pap. cclvi 
verso) ; but these exhaust the list of literary texts contain- 
ing such symbols. In non-literary papyri, as might be 
expected, they are more freely used, especially in accounts 
and receipts. Still more common, however, than the use 
of symbols to denote the terminations of words [k, t, and 
the like) is the practice of abbreviating words by the 
simple omission of terminations. This is found in tiie 
Aristotle and medical MSS. above-mentioned, but is es- 
pecially frequent in non-literary documents. The regular 
system of such abbreviations is to omit the latter part 
of the word, and to elevate the last letter remaining above 
the line, or else to draw a line over it as a mark of 
abbreviation ; thus either -rrpo" or irpoK may stand for 
TTpoKeiTau Abbreviations such as these explain themselves, 
and do not admit of tabulation ; but a list of symbols used 
in abbreviations is given in an appendix. Contraction, 
in the sense of the omission of the middle portion of 
words, such as occurs in mediaeval Latin MSS and in 
modern letters, is not found in Greek papyri ^. 

The study of tachygraphy is too special a subject to Tacliy- 
be dealt with here, and the explanation of the few extant ^^'^^^ ^■^' 
specimens of it on papyrus is still obscure ^. No long 
document in this style of writing has yet been discovered ; 
and though several small examples are said to be in the 
Rainer collection at Vienna, they have hitherto been only 
imperfectly published, and the explanation of their systems 
has, in most cases, still to be given. 

'■ Scott, Fragmenta Herculanensia, p. 98. 

'^ A supposed instance to the contrary has been pointed out in Grenfell's 
Greek Papyri, i. 24 (now Brit. Mus. Pap. Dcxx), 1. 6, where ^aaar)s is given 
as the reading of the MS. for ^aoLXiaa-qs. In reality the word is not con- 
tracted at all, biit simply written in a very cursive fashion. 

^ See (in addition to the older literature) Gitlbnuer, Die Drel Syshme der 
griechischen Tachygraphie (1894), and W^essely, Bin System altgrieddscher Tacliy- 
graphie (1895), both in the Benkschriflen of the Vienna Academy. 




The three The history of Greek writing upon papyrus has three 
papvnjs well-marked periods, the distinction between which is the 
paiaeo- foundation of all palaeographical knowledge of the subject, 
.r^rap ly. 'j'ljggg periods correspond to the three political administra- 
tions by which the country was successively governed after 
the extinction of the native Egyptian empire. From 323 
to 30 B. c. it was under the sway of the Ptolemies ; from 
the conquest by Augustus to tlie re-organization of the 
empire by Diocletian, it was administered from Rome ; 
and from the time of Diocletian to that of the Arab 
conquest in A. D. 640, it was a part of the eastern or 
Byzantine division of the Roman world. The changes in 
the prevalent type of handwriting curiously reflect the 
changes in the administration ; and the classification of 
hands as Ptolemaic, Roman, or Byzantine is not merely 
conventional, but corresponds to real differences of character 
which can be made obvious to the most untrained eye. 

It is, however, only of the non-literary hands that this 
proposition is true in so extreme a form ; and the reason 
for this does not seem difiicult to imagine. The pattern 

^ In parts of this chapter I have made use of my own treatment of the 
same subject in the Introduction to the first volume of the Catalogue of 
Greek Papyri in the British Museum (1893), but with amplifications and modi- 
fications according to the new material which has come to light since 
that was written. The book is, of course, one which will have come into 
the hands of few but specialists. 


for the non-literaiy hands — not merely for those of pro- 
fessional clerks, but, through them, also for those of private 
persons — was set by the government officials, and varied 
according as the higher members of that class came from 
Alexandria, Rome, or Constantinople. The literary class, 
on the other hand, had no direct dependence on the 
political capital. If they were dependent on any outside 
influence, it was that of Greece, the fountain from which 
their higher inspirations were drawn, and with which they 
were connected by a strong and continuous tradition. Only 
gradually, and at some considerable distance, were the 
fashions of literary manuscripts affected by the contem- 
porary varieties in every-day writing ; and it requires 
a little practice to see where the characteristics of non- 
literary papyri manifest themselves in the literary hands 
of the same period. That they do manifest themselves, 
however, will, it is hoped, be shown in the following 
chapters; and the full and certain knowledge which we 
now have of non-literary palaeography goes far to lay 
a firm foundation for the more interesting and important 
branch of the subject which deals with the literary 

The beginning of the history of papyrus-palaeography I- Ptole- 


is fixed, for the present, by the discoveries of Mr. Flinders period : 
Petrie in 1889, when he extricated a mass of documents of ^^^^ Petrie 
the third century b. C. from a number of mummy-cases found 
at Gurob. The mummy-cases, instead of being of wood, 
were made of a kind of papier-mach(^, the material being 
papyrus, coated over with plaster. The papyri were evi- 
dently nothing but the produce of the waste-paper baskets 
of the period, torn, cut, pasted together just as they came, 
defaced by the plaster and mutilated by rough handling ; yet 
these same rubbish-heaps, patiently sorted and set in order 
by Prof. Mahaffy, are now the foundation of our knowledge 
of Greek palaeography. To these must be added the great 
Revenue Papyrus of Ptolemy Philadelphus, the greater 

D % 


part of which was acquired by Mr. Petrie in 1^93, and the 
rest by Mr. Grenfell in the following year ; but this only 
confirmed by additional examples the knowledge which 
the Petrie papyri had already established. Before Mr. 
Petrie's discovery there were, it is true, a few documents 
of the third century already extant in some of the great 
European libraries ; but they were undated, and there was 
nothing to show their real age, which, in consequence, 
was generally underrated. About the age of the Petrie 
papyri there could be no doubt. Many of them were wills, 
petitions, and similar documents, bearing precise and 
indisputable dates in the reigns of the earliest Ptolemies 
(with the exception of Ptolemy Soter, the founder of the 
dynasty). The earliest definite date was in the year 
270 B.C., the latest w^as in 186 B.C.; and there was no 
reasonable doubt that the great mass of undated documents 
lay between these limits, and that nearly all of them 
belonged to the third century B. c. Subsequent examina- 
tion of the previously undeciphered fragments of the collec- 
tion has, indeed, revealed one or two later dates, but has done 
nothinof whatever to shake the g-eneral conclusion. From 
some seventy definitely dated documents in a great variety 
of hands, and from many scores of undated documents 
of the same period, we have ample means for estimating 
the character of Greek writing in Egypt — and specifically 
in the Fayyum — in the third century b. c.^ 

The first characteristic which strikes the eye in the 
writing of this period as a whole is its freedom and 
breadth. The style is light and flowing, strokes are free 
and curved, without being necessarily careless and ill- 

1 The descriptions which follow will be more intelligible if read with 
the series of facsimiles published by the Palaeographical Society, or the 
atlases accompanying the Cakilogue of Greek Papyri in the British Musenm (vols. 
i. and ii.\ A very useful table of alphabets is given in Sir E. M. Thomp- 
son's Handbook of Greek and Latin Palaeography. The facsimiles given in the 
present volume, though fairly characteristic of their respective periods, are 
too few to exhaust all their varieties. 

Plate I. 



! J - 














'^ >-• T- V' '-tL 



1 1 



■■• J. 


•s •• 

"- r 


formed. Such hands, indeed, there are among them, in 
which cursiveness borders closely on illegibility; but the 
characteristic hand of the period is graceful and easy, 
showing at once a full command of the pen and a plentiful 
supply of papyrus. Letters such as M, TT, T have an almost 
excessive breadth in their horizontal strokes, and it needs 
a good-sized piece of papyrus to contain any appreciable 
quantity of writing. The Roman style is altogether smaller 
and more compact; that of the Byzantine period, though 
its letters are sometimes quite as large, is squarer, gene- 
rally heavier, and shows more signs of deliberate care 
and conscious style. The Ptolemaic scribe wrote freely 
and often well, but without self-consciousness. 

Another characteristic of Ptolemaic writing is the appear- 
ance, which it generally presents, of a horizontal line along 
the top of the letters. Most of the letters are shallow, and 
the horizontal strokes in them are formed near the top, 
while the perpendicular strokes project very little above 
the line and are carried far down below it. Thus M (a very 
characteristic letter in Ptolemaic hands) is generally formed 
of two perpendicular strokes, stretching well below the 
line, and united at their tops by a horizontal stroke with 
a very shallow curve. This feature is not found in either 
of the later periods, and indeed is more characteristic 
of the third century B. c. than of the later Ptolemaic age. 
Good writing of the Roman period has few prominently 
projecting letters at all ; while the Byzantine hand, though 
marked by very long upright strokes, presents the appear- 
ance of an even line along the bottom of the writing, with 
projections bristling along the top. 

The specimen shown in the accompanying facsimile 
(Plate I) is part of a letter of the year 242 B.C., and is 
a fairly characteristic example of the hand of this period. 
The completely cursive character of the writing is obvious 
at first sight, as well as the line along the tops of the 
letters described above. There are no signs of difficulty 



The third 
B. c. 

B. c. 

or of unfamiliarity with the use of the pen. It is clear 
that writing of such freedom and ease has a long history 
behind it. 

With reofard to the forms of individual letters in docu- 
ments of the third century, the most characteristic are <^, 
M, N, TT, T, Y, and CO. ^ is often found, especially in the 
less carefully written papyri, in the shape of a simple wedge 
{/^,/-)] M, as described above, has a very shallow curve 
between two nearly perpendicular strokes ( n) ; N frequently 
has its final upright stroke carried up far above the line 
(y-v, r*) ; TT is either broad and low or rounded into an almost 
semi-circular curve ; T is almost always without the right- 
hand portion of its cross-bar, being written with a single 
stroke of the pen ; Y generally has a large loop, carried 
further to the left than the right (^) ; while CO habitually 
has its second loop represented by an almost or quite 
straight line. These characteristics are, no doubt, most 
apparent when the writing is least careful and formal ; but 
few documents of the period are without some of them. 
There is very little modulation of fine and thick strokes. 
As a rule the lines, though not heavy, are thick and black ; 
and there is much unevenness in the size of the letters. 
Hence the writing of this century, free and flowing though 
it often is, is rarely very ornamental. The enthusiasm 
of the first discoverers led them at times to speak too 
highly in its praise ; but an unprejudiced comparison of it 
with the hands of the succeeding centuries will show that, 
at least in the specimens hitherto known, it is inferior 
in regularity and handsomeness. 

For the first half of the second century b. c. our material 
is mainly drawn from the papyri found in the Serapeum 
at Memphis, many of which were written by a single scribe, 
a Greek recluse in the Serapeum named Ptolemy. There 
are, however, several examples of other hands, which justify 
us in using this group of documents as fair evidence for 
their period. As a whole, these hands are larger, more 

Plate II. 

''^WWW»-«^yptlt^ftHKipymy,<^li |, l ^^j ip „ ..^vJT-.«> WB i ^Mm i,,», ., ■•m y^ , 

,; r.f|, 


-*5 It^-f 

- a 



regular, and less cursive than those of the preceding cen- 
tury. The letters have more of an uncial form, and are 
only linked together in groups of two and three. The 
better-written specimens are clear and regular, and some- 
times even handsome. The appearance of a line along the 
top of the writing ceases almost entirely. Of the individual 
letters, the wedge-shaped A is rarely found after the third 
century ; it is of the minuscule type, but the loop is some- 
times represented by a mere straight line (A") ; M is still 
sometimes marked by its shallow depression, but oftener 
the middle loop is carried lower down and bent to an angle 
in the centre, so as to resemble a rough uncial M; B is 
very large, often extending both above and below the line ; 
the first stroke of H is higher than the second, which 
rarely rises above the cross-bar, and is linked to the 
following letter by a horizontal stroke from the top ; 
K is generally large, especially the upright stroke ; N is 
sometimes of the third century pattern, with the last stroke 
rising above the line, but is oftener of the normal uncial 
shape ; T has acquired the right-hand portion of its cross- 
bar, though it is still often written without lifting the pen, 
by making the cross-bar first and then drawing the pen 
backwards and downwards; at other times the first half 
of the cross-bar and the down-stroke are formed together, 
as in the third century, and the second half of the cross-bar 
is added separately, being often attached to the succeeding 
letter ; Y is not unlike T, being made by forming a wide, 
shallow curve, and then drawing the pen backwards and 
downwards ; (x) is generally of the ordinary minuscule 
type, though the second loop is still sometimes clipped. 

The facsimile shown in Plate II (from Brit. Mus. Pap. 
XLiv) is, perhaps, the best specimen of calligraphy among the 
Serapeum papyri, though another (Pap. xxiv) is in a larger 
and bolder hand. It is a petition from the above-mentioned 
Ptolemy, son of Glaucias, for protection and redress on 
account of an assault made upon him by some of the 


Egyptian attendants in the temple. It is dated in the year 
i6i B.C., and is written in a clear, regular, and notably 
well-formed hand — certainly not Ptolemy's own, but 
probably that of a professional scribe. 
Late se- ^he second half of this century was almost a blank, so 

cond cen- "^ ' 

turyB. c. far as palaeography was concerned, until the purchase 
by Messrs. Grenfell and Hogarth, in 1894-5 and 1895-6, 
of a large number of dated documents in very good 
condition, which cover this period very fully and extend 
into the following century. In these papyri, which come 
from the neighbourhood of Thebes, we see the revival 
of a fully cursive style of writing; or perhaps it would 
be more true to say that they justify the belief that a more 
cursive style than that of the Serapeum documents was in 
existence during the earlier half of the century, as it had 
been in the third century. The cursive hand of 150-100 b. c. 
(see Plate III) is, however, quite distinct from that of 270- 
200 B. c. It is smaller, better formed, and more ornamental ; 
with quite as much liberty as its predecessor, but less 
licence. The scribes have learnt to be regular without 
being stiff, and the sizes of the letters are Ijetter propor- 
tioned to one another. It is the best period of the 
Ptolemaic cursive, and comes just before its decline and 
disappearance. At the same time, it is the most difficult 
style of writing to describe in words. The forms of the 
individual letters are less noticeable and peculiar than 
in the earlier hands, and approximate to those of the 
Roman period ; and yet the general aspect of the writing 
is unmistakably Ptolemaic. It lacks the roundness of the 
Roman style, and the letters are of a uniform thickness, 
without modulation of broad and thin strokes, with a ten- 
dency to thickness and blackness throughout. Of the 
individual letters, A is generally small, and its loop becomes 
either a round spot or a straight line; H generally has 
a ligature attached to its last stroke, whether there is 
another letter for it to be linked with or not; the same 



Plate III 




LOAN. — B.C. 105. 


is the case with N ; Y fluctuates between its earlier shape 
and one more resembling a Y, of which the left-hand arm 
is generally longer than the right. But, on the whole, 
little reliance can be placed in the forms of single letters at 
this period ; while, on the other hand, the general appear- 
ance of this small, even cursive can hardly be mistaken. 

The first century was, until quite lately, the most obscure The first 
pei'iod in the whole history of papyrus-palaeography; and ^c. 
it cannot even yet be said to be adequately known. On 
the one hand there were a few documents dated between 
100 and 80 B.C., which carried on the tradition of the 
preceding century without much recognizable variation ; and 
on the other there were a few which belonged to the last 
decade of the century, in which the Roman cast of hand was 
already well developed. The interval of transition is now 
precariously bridged by some papyri acquired by Messrs. 
Hogarth and Grenfell during their campaign on behalf 
of the Egypt Exploration Fund in 1895-6. The chief 
conclusion to be derived from these is that the transition 
was very rapid. In the early part of the century the 
Ptolemaic cursive is seen to be breaking up. It becomes 
less firm and regular, and loses its sense of style. Under 
Ptolemy Lathyrus and Auletes it is an ugly and a broken 
hand. In the middle of the century (if the dates assigned 
to these documents, which are sometimes uncertain, are to 
be trusted), several forms of letters which are characteristic 
of the reign of Augustus are found intermixed with 
Ptolemaic forms ; and by the close of the century the 
Ptolemaic forms have entirely disappeared, and the writing 
is distinctly Roman. The only criterion that can be given 
for the determination of MSS. of this period is to remember 
that it is an age of transition, and to look for an inter- 
mixture of forms. The styles of the end of the second 
century b. c. and of the beginning of the first Christian 
century must be known first, and the intermediate stage 
will then be recognizable. 


11. Roman If* one special characteristic is to Ije assigned as, more 
^^^^ ' tiian any other, marking off the Roman period from its 
predecessor and its successor, it is roundness. Straight, 
stiff lines are avoided ; curved, flowing strokes take their 
place wherever possible. Ligatures, which in the Ptolemaic 
cursive hand were generally straight lines, are generally 
curved in the Roman hand. There is more modulation 
of strokes, and the somewhat thick and black aspect 
of Ptolemaic writing disappears. It also becomes usual 
to write letters as much as possil)le without raising the 
pen. This is especially noticeable in the case of 6, the 
cross-bar of which is now very commonly attached to its 
upper curve (C", C); and this formation may be taken 
as an almost conclusive proof of a Roman date. In early 
Roman documents, on either side of the beginning of the 
Christian era, Y is very noticeable for the deep curve 
of its upper part and rather prominent development of its 
tail (1/ ). Other letters besides € exemplify the tendency 
to form letters in single strokes. The oblique strokes of K 
become a curve attached to the bottom of the upright 
stroke by an obliquely-rising ligature, so that the whole 
letter assumes a shape resembling a written u. An almost 
identical shape is often assumed by B, which begins in this 
period to be frequently formed with open top, though the 
capital form continues contemporaneously. The cross-bar 
of N becomes a curve uniting the tops of the two upright 
strokes (/X). A change is also observable in T, which 
becomes fork-shaped in many instances, and sometimes 
almost has the form of a V (T, y). Even cj) is written 
without raising the pen, being formed of a semi-circular 
curve, the end of which is attached to the top of the 
perpendicular stroke, and the latter not unfrequently falls 
outside the curve altogether; at other times the circle 
is represented by a stroke resembling an s lying on its 
side, through which the perpendicular stroke descends. 
These characteristics run, more or less, through all the 

Plate IV. 









'h -^- in !''^«i At" -'d .-^^ r- S- x^-§- 

2 -?-■>  -^'^ii ■^■1 . ■^ '^■*v^ '^ — * ,C' Cl i  

*?' rv _. 
A-'f^. '.k:.*"-^^ ; "Ji"" V .-t r^ ^ '»' 'X ^ ^ 

JS., -.AT' -i... X r\r 





^;-a >':'^.^-^' 



^fsT .X ft ,'\ 


-<i.^:ik ^''««i.'iutii' 

Plate V. 




















*"x: * 






• <-V 


V :> 














. 3 



i '^ 

r^ ■*':'' ; 





Roman period, serving to differentiate it from the Ptolemaic First 
and Byzantine ages, but not sufficing for the accurate *^*''" '">'• 
dating of MSS. within the first three centuries. This is 
a matter partly of general appearance, partly of variations 
in a few individual letters. Documents of the reign of 
Augustus and his successors as far as the middle of the 
first century are generally written in a bolder and more 
angular hand than those of a later date ^ The forms 
of the letters are more pronounced, as in the case of Y, 
mentioned above, € and K. In the middle of the century 
these angles have been smoothed away, and the round, 
graceful character of the best Roman cursive is fully 
developed. Indeed the hand may be said to be at its best 
between the years 50 and 100. The specimen here given 
(Plate V), from a poll-tax register of A. D. 72-3 (Brit. Mus. 
Pap. CCLx), is a good example of the larger hand of the 
period, and one which will be found useful in assigning 
a date to some literary papyri in a later chapter. At the 
same time a smaller hand, and one much more difficult 
to read, came into fashion in the reign of Domitian ; and 
small hands, though not unknown earlier in the Roman 
period, become predominant from this point until the end 
of the second century or later ; increased cursiveness going, 
as usual, with reduced size of the letters. An example, 
from the extreme end of the first century, is given in 
Plate VI (Brit. Mus. Pap. cxliii). 

Two letters are especially useful as indications of manu- H and C. 
scripts written in the latter part of the first century and 
the first half of the second. These are H and C. The 
former letter, during this period, has two quite distinct 
forms. One is the familiar H-shape, in which -the cross-bar 
usually rises slightly and the second upright descends from 
it in a shallow curve ; this is found more or less throughout 
the period, and in itself contains no criterion of age. But 

1 Plate IV, an official document from the beginning of the reign of Tiberius 
(Brit. Mus. Pap. cclxxvi), is a fairly characteristic example of this style. 


by the side of this there is another form, consisting of a 
horizontal curve with an ahnost perpendicular tail (^, "|), 
which is characteristic of a comparatively limited space 
of time. In appearance it resembles the Y of the same 
period, and is often only distinguishable from it by the 
fact that its tail is either perpendicular or curls to the left, 
while the tail of the Y curls to the right. This form of 
H is, in point of origin, a modification of the ordinary 
H-shape, in which the first perpendicular stroke is so 
much slurred as almost to disappear, and accidental ap- 
proximations to it are occasionally found in much earlier 
documents ; but to these no importance can be attached, 
and the real range of its currency may be stated as being 
from about a.d. 50 to about i6o^ There are sporadic 
instances later than this, as is inevitably the case with 
every palaeographical characteristic ; but the material for 
the first three centuries is now so plentiful that we are 
justified in asserting that the practical disappearance of 
this form at about the date named is not accidental, 
or a fancy based on the absence of evidence, but really 
represents a fact in palaeography. 

In the case of C the data are less precise, but still useful. 
The C of the Ptolemaic period is upright and has invari- 
ably a horizontal top (c )• In the early Roman period the 
upright form is still maintained, and often the horizontal 
top as well, though the Roman fondness for curves shows 
itself in the increasing use of a rounded top. Towards 
the end of the century the curve has a tendency to be 
carried further over, and the letter assumes a tumble-down 
appearance [C ,C,^); while in the next century it sometimes 
even passes the horizontal stage, and becomes a backward- 

' It is found very plentifully in the first and fourth hands of the 
Aristotle papyrus (indeed it was here that its existence was first noted), 
written about a. d. ioo, and also in the accounts on the redo of the same 
jjapyrus, which were written in a. d. 78-9. Plate VI, a receipt of a. d. 97, 
not only has this letter but generally resembles the style of the Aristotle 
MS., and confirms the date assigned to it. 

Plate VI. 

4. .:f.):^VTAv>/^Ct-»'■^^^^^r^i^^^^ -^T/^-Kmc^o^^- 

RECEIPT. — A.D. 97. 

Plate VII. 


i ♦- 


RECEIPT. — A.D. l66. 


facing curve ('^, O; e.g. <^^ — a?). The upright C witli 
flattened top continues simultaneously with this form 
throughout the whole period, and consequently is not 
to be taken as a sign of early date ; but the tumble- 
down C, according to the fairly plentiful evidence which 
we now have on the subject, can hardly be earlier than 
the end of the first century. Early in the third century 
the upright form seems to have been universally resumed, 
though in a larger and coarser type. 

For the second century palaeographical material is more Second 
plentiful than for any other period in the whole history ^®" ^"^" 
of writing upon papyrus, not a single year being un- 
represented by at least one accurately dated document. 
It is, of course, useless to pretend to lay down precise 
laws for the discernment of documents of successive decades, 
but the general lines of development may be indicated. 
In papyri of the reigns of Trajan and Hadrian, the 
letters are usually of fair breadth, and are reasonably 
well formed ; while in the reign of Antoninus Pius, and 
still more of Marcus Aurelius, the increased cursiveness 
of the prevalent hand leads to the compression of letters 
and often to their slovenly formation. A very small 
and very cursive hand is especially characteristic of the 
reign of Marcus ^ ; and in many of the receipts and leases 
of this period it requires little less than divination to make 
out the scribe's intention — especially if, as often happens, 
illiteracy be added to cursiveness. A reaction from this 
very minute style is manifest in the reign of Commodus 
and under the emperors of the first part of the third 
century; but though the writing gets larger, there is no 
regeneration in the formation of the letters. On the con- 
trary, from this point may be dated the break-up of the 
Roman hand. The writing becomes rough and coarse, and 
often extremely ugly. Letters are less formed and worse Third 
formed, and the writing straggles unevenly over the page. 

* See Plate VII, a receipt of a. d. 166 (Brit. Mas. Pap. ccjxxxii). 


To this rule, as to all others, there are exceptions ; one 
(Brit. Mus. Pap. cccxxii) so marked as to make us dis- 
trust the cogency of the argument which, on the ground 
of the occurrence in it of the name of Aurelius, assigns it 
to this period ^ 

About the second quarter of the third century a marked 
improvement takes place, for which no adequate reason can 
be assigned, unless it be a mere accident in the survival 
of evidence. As a rule, the variations in handwriting in 
Egypt curiously correspond to the changes of government. 
The rise of a new government is accompanied by the ap- 
pearance of a new style of writing, and the decay of the 
writing goes hand in hand with that of the administration. 
The rise of the Ptolemies, the collapse under Auletes and 
his fleeting successors, the advent of the Eomans and 
the firm rule of the early emperors, the decline when 
Commodus and Caracalla succeeded Antoninus and 
Marcus, the reorganization of the empire under Diocle- 
tian, all have their palaeographical parallels in the docu- 
ments of the period ; but there seems to be no reason 
why the chaotic years about the middle of the third 
century should show any improvement on their prede- 
cessors. The fact, however, remains visible in such 
evidence as we possess, and there are several extant 
documents between a.d. 250 and 280 which are written 
in tidy and even ornamental hands [e.g. Brit. Mus. Pap. 
CCCLI, Paris Pap. 69 e, and a Berlin papyrus shown in 
plate xiv of Wilcken's Tafeln). 

At this point in the history (about the middle of the 
third century), the evidence which has been so plentiful 
since the beginning of the Christian era suddenly becomes 
extremely meagre. Indeed for the important period of 
transition in the reign of Diocletian evidence was wholly 

' The use of the name Aurelius by private individuals was granted in 
A. D. 2T2, and thenceforward it becomes so common as almost to be uni- 
versal ; but there are certain instances of its use as far back as a. d. 175. 


wanting until the acquisition by Mr. Grenfell, in 1895-6, 
of a small group of documents (now in the British Museum) 
from the years on either side of the turn of the century. 
For the present, therefore, one can only speak with some 
uncertainty as to the details of the change. That a change, 
however, did take place, is manifest on a comparison of 
papyri of the middle of the fourth century with those of 
a hundred years earlier, the difference between them being 
quite as marked as that between papyri of the reigns of 
Auletes and Augustus. The transition period is brief, but 
has a well-marked character of its own. Writing becomes 
smaller again, though not so small as in the second century. 
It is compressed laterally, and the letters are stifFer and more 
angular in formation, with a tendency to lean forward. 
It looks as if the formalism of the Byzantine age began 
to impress itself upon its scribes even from the very 
beginning of the revolution brought about by Diocletian. 

For the fourth century nearly all the evidence at present Fourth 
extant consists of a single group of papers, containing ^^^ '^^^' 
the correspondence of a certain Abinnaeus, who was the 
commander of a body of auxiliary cavalry quartered at 
Dionysias, in the Fayyum, for some years on either side 
of A.D. 350. The larger part of this collection is now 
in the British Museum ^ ; the rest are at Geneva. To 
these have to be added two examples of the more orna- 
mental writing of the period, published in Wilcken's 
Tafeln (Nos. xv. and xvi.), and a few magical and other 
undated papyri which, in the light of the Abinnaeus 
papyri, we can now confidently assign to this century. 
There are also some good specimens in the Rainer collec- 
tion at Vienna, but these are not yet published; and 
the same is the case (so far as facsimiles are concerned) 
with the fourth-century documents among the Oxyrhynchus 

' Many of these are reproduced in the atlas of facsimiles accompanying 
vol. ii. of Greek Papyri in the British Museum ; and one, which is a good 
example of tlie better hands in the collection, is given in Plate VHI. 


papyri and at Berlin. The most striking general charac- 
teristic of fourth-century writing is an increase of size and 
(in the better written and representative examples) of calli- 
graphic style and ornament. The letters are boldly formed, 
with sharp angles and prominent strokes. The better hands 
are generally composed of fine and delicate lines, which 
emphasize the sharpness of the angles and the boldness 
of the curves; but the more illiterate hands, of which 
there are many among the Abinnaeus papyri, are often 
very thick and coarse. Of single letters the most notice- 
able are K and O, on account of their increased size and 
prominence. The sharp angles and projecting upper 
strokes of K almost always catch the eye in documents 
of this period ; while O, which in Ptolemaic and Roman 
writings is often reduced almost to a dot, is now a large 
and conspicuous letter. The right-hand stroke of A is 
often widely separated from the rest of the letter; and 
sometimes a d-shaped letter takes the place of the familiar 
form. The upper half of e is a straight oblique stroke, 
pointing to the right {(:) ; while rj has already acquired 
the h-shape which belongs to it throughout the Byzantine 
period. C once more stands upright, generally with a 
flattened top, as in the Ptolemaic and early Roman age. 
Finally Y is v-shaped, and ceases to be a conspicuous 
letter. The general impression left by a page of good 
writing of this period is one of regularity and order, 
though not so mechanical as it subsequently became. 
III. By- The largest unexplored tract now left in the history 
Period^ of cursive writing on papyrus is that from about A.D. 
Fifth ^5q Iq about A.D. 5CO. A few documents from the 

cen uiy. q^,^^^ Oasis, recently acquired by the British Museum from 
Mr. Grenfell, represent the years on either side of 400, and 
tend to show that the style current in the middle of the 
century had not greatly changed by then. The Rainer col- 
lection possesses some fragments of the fifth century, but 
these have not been published, and are not available for 

Plate VIII. 







f • 




/■■U— ; 

^{,k^i~o "^ " 

'-■•^k^'i^^lm <^M >frf' <^-vo 

LETTER. — C/rC. A.D. 35O. 


comparison. It is not until close to the end of the century, 
in A. D. 487 and 498, that dated materials are again acces- 
sible, and these cannot be separated from the sixth century 
in general. There, on the contrary, material is once more 
plentiful. The first Fayyum find, which has supplied 
documents in thousands to Vienna and Berlin, and in 
lesser quantities to Paris, Oxford, and London, consisted 
almost entirely of papyri belonging to this later Byzantine 
period, extending from the beginning of the sixth century 
to the end of Greek writing in Egypt. Unfortunately, 
the number of precisely dated documents of this period 
bears a very small proportion to that of the undated 
or insufficiently dated ^ This is largely due to the 
practice of dating by the fifteen-year period known as the 
indiction (see below, p. 54), which by itself is quite useless 
for the purpose of determining the age of a document 
after the lapse of a few years. The consequence is that, 
while it is generally very easy to assign a hand to the later 
Byzantine period, it is very difficult to determine whether 
it is of the sixth or seventh century. The difficulty is 
increased by the conservatism of Byzantine scribes, who 
adhere for a considerable time to the same type of hand. 

Speaking generally, the fully-formed Byzantine hand Sixth and 
is a large, well marked, and rather handsome hand ; not centuries, 
so delicate as the best examples of the fourth century, 
but regular, with ornamental strokes and curves, and 
with an unmistakable air of formality. In the best 
specimens the writing is upright and square, with plenty 
of width, so that any given piece of Byzantine papyrus 
contains much less writing than a similar piece of the 
Roman, or even of the Ptolemaic period. Many of the 
letters, also, are formed in quite peculiar manners. The 

' This defect will be rectified when dated facsimiles of the Oxyrhynchus 
papyri of the Egypt Exploration Fund are published. The si^ecimen 
shown in Plate IX is not dated, but seems from its style to belong to the 
sixth century. 



loop of a is large and generally open at the top, resembling 
a modern a more than any a of the earlier periods. /3 is 
generally a long, irregular oval, open at the top, antl 
with a short tail projecting downwards from the bottom 
right-hand corner. The minuscule form of 8, resembling 
a Latin d, is increasingly common, and has a very long 
upright stroke; but the uncial form is also found. The 
upper half of e is still more pronounced than in the fourth 
century, projecting obliquely for a considerable distance 
above the line {4). rj is generally h-shaped, with a very 
long upright stroke, like 8. Similarly, the first stroke 
of AC projects far above the line, the rest of the letter 
being u-shaped. The left-hand stroke of A comes far 
below the line, and is often widely separated from the 
right-hand stroke, /x is in its minuscule form, much as 
it is here printed, with its first stroke perpendicular 
and stretching far below the line, u varies between the 
uncial type, in which case the junction of the oblique 
stroke with the second upright often takes the form of 
a curve ( U , kl), and a cursive form resembling a Latin n. 
or is well rounded, generally divided in the middle, with 
the upper half sometimes rather exaggerated, r, if not 
of the ordinary shape, has a long tail and is deeply 
forked at the top. v is small and v-shaped, and not 
unfrequently becomes little more than a curved ligature 
in combination with other letters. The ujDright stroke 
of is generally united to the circle by a well-rounded 
curve, and the whole letter is usually large and prominent, 
— as, it may be remarked, it likewise is in vellum uncials 
of the same period. 

Besides this upright hand, a sloping style is also found 
in use during these centuries. A papyrus in the British 
Museum (cxiii. 5 6) shows it in existence in A. D. 542, but 
it is especially characteristic of the later years of this 
century and of the seventh century. The shapes of the 
individual letters are much the same as those just described, 

Plate IX. 





but they assume a marked slope to the right, become 
smaller and less square, and are laterally compressed. 
a is more closely connected with the letter which follows 
it, and is often a mere loop, like the v described above. 
The upper half of e is more rounded, and, with the central 
cross-stroke, is often separated from the lower half. 
The left-hand stroke of A is still more prolonged, con- 
tributing much (like the up-strokes of 8 and r;) to the 
sloping appearance of the writing. The tail of the first 
stroke of fi is shortened, v is not unfrequently written 
above the line, in the shape either of a very wide and 
shallow curve, or sometimes of a straight line. 

This sloping hand is especially characteristic of the 
later part of the Byzantine period, but it did not ex- 
tinguish the upright style of writing, which is found 
even in the seventh century. For instance, the British 
Museum has a good example of it dated in the year 619 
(Pap. ccx). Hence it is unusually difficult to assign a 
precise date to any undated document of this upright type, 
and one must commonly be content to define it merely 
as ' sixth or seventh century ' ; though in the later examples 
the characters are generally less firm and precise than in 
the earlier. 

The two types of hand which have just been described 
are, both of them, large, and cover a good deal of ground ; 
but there is also a small type of hand, which is used 
chiefly for accounts and receipts. These are usually 
assigned to the seventh, and sometimes to the eighth, 
century ; but their age is really very uncertain, since 
few are dated in any other way than by indietions. 
The forms of the letters are much the same as those of 
the larger hands, and the projecting strokes of such 
letters as 8, 77, A, /x, are equally marked ; but their smaller 
size calls attention to the fact that the Byzantine hand, 
at any rate in its later stages, is definitely a minuscule 
hand. Whether written large or small, the forms of 

E 2 


I nearly call the test-letters are now minuscule ; and the 
special interest of these small hands is that they indicate 
the way to the transition from the minuscules of the 
papyri to those of the vellum MSS. of the ninth and 
tenth centuries. More will liave to be said on this point 
in a later chapter ; for the present it must suffice to state 
the proposition that, while the literary hand of the Roman 
period is the parent of the vellum uncial, the Byzantine 
* non-literary hand is the parent of the -vellum minuscule, 

' which comes to the front in the ninth century as a book- 

hand for the first time, and establishes its supremacy 
in the tenth. 
Tlie end of The exact end of Greek writing upon papyrus cannot 
writing on he fixed; but it is certain that the Arab conquest of 
papyrus. Egypt in A.D. 640 gave it its death-blow. Documents 
that can be placed with certainty later than this date 
are rare ; probably there are more at Vienna than 
anywhere else, but they are not yet accessible. The 
latest papyrus of any length with a precise date is of the 
year 683 ; but one long document in the British Museum 
(Pap. lxxvtt) is probably of still later date than this'. This 
is the will of Abraham, bishop of Hermonthis and head 
of the monastery of St. Phoeliammon at Djeme, near 
Thebes. It forms one of a group of documents, the rest 
of which are written in Coptic ; one of these is dated 
in the year 786, and several others are shown by internal 
evidence to belong to about the same date. The will 
of Abraham appears to come near the beginning of the 
series, but there is nothing to show that it falls outside 
the eighth century. It is a large and fairly upright 
hand, showing that the traditions of the sixth century 
lasted on even till this late date ; only the somewhat 
broken and degenerate look of the writing, the roundness 
and looseness of the shapes of the letters, distinguish 

' Complete facsimile in atlas accompanying vol. i. of Greek Paprjri in the 
British Museum. 


it from the square and precise appearance of the hand 
from which it is descended. It marks, however, the close 
of Greek writing in Egypt ; and it is noticeable that the 
bishop, whose testament it is, is expressly said in it to 
be ignorant of the Greek language. The Greek language 
was, in fact, gradually extinguished by the Arab conquest ; 
and with it disappears our knowledge of Greek writing 
on papyrus, since in no other land than Egypt has the 
brittle material survived to our own time. The full 
history of the transition from papyrus to vellum can 
never be written, for want of the materials. 

One detail of importance in connexion with non-literary 'i'}i« tlatmi 

. . . . ut piipyri. 

papyri must be mentioned before closing this sketch of 
their history. This is the manner in which they are 
dated. The formulas of dating differ, like the hands 
themselves, in the Ptolemaic, Roman, and Byzantine 
periods. In a Ptolemaic document the date is given by 
the regnal year of the reigning sovereign, and the full 
formula for this contains not only the name of the king 
himself, but also a list of the priesthoods of all the defunct 
Ptolemies ; for example (Brit. Mus. Pap. Dcxxiii). 

^acrik^vovToiv KktoTrarpas kol /SacrtAews nro\eju.aiou Oeo^v 
<t>L)\.oixrjT6p(i>v ^(ji)Ti]p(x)V irovs ?], e<^' tepecos rod ovtos (.v 'AAe^- 
avhpda ^ AXe^dvbpov Kal decav ^cbTi]poov koI deu)V 'Abek(f)Qv 
KOI 6t&v Evepy€TO}V kol Q^Gjv ^iko-naroputv koX 6(.S>v ^Ein(j)avu>v 
Kol S^ov EvTrdropos kol O^ov ^Lkop.i]Topoi koI Oeov ^ikoTtaTopos 
viov Kttl d€OV Evepyirov kol Oecov ^ikopirjTopMV '2a>Ti]pot)V, 
lepouTTCoAou *'l(n8os /xeydA?js //rjrpo? O^&v, ddko(p6pov Bepei't/cr^s 
Ev€pyeTtbos, K.avr)cf)6pov 'ApaLPOi-jS 4>tAa8eA(^ou, Upeias ^ApaLv6y]s 
<t>tAo7raropos', rwy ov(TU>v kv 'Ake^avbpeCq, ev be Ylrokep-albi t7]s 
0?)/3ai6oj €(/)' Upeoov koL tepetwf kqI Kavrjipopov T(av ovrcav /cat 
ovaSiV, pLTjids Mex^tp Td iv KpoKobCkcov TTo'Aei tov HadvpiTov. 
All this merely describes the year which we indicate as 
B.C. 109. But since such a formula might be accused 
of being cumbrous, and of including a good deal of un- 


important matter, it is very often reduced to tlie simple 
statement hovs 7\, M^xdp Td : which errs as much from 
brevity as the other does from length, since it omits 
the name of the king, and it is not always that the 
reign can be identified with certainty. 

Roman dates are more business-like, but they too have 
a longer and a shorter formula. Dates are given by the 
regnal year of the emperor; but in both Ptolemaic and 
Roman dates it must be remembered that the year always 
begins with the ist Thoth (—29th August). Thus the 
first year of a sovereign lasted only from his accession 
to the 1st Thoth next ensuing. The longer Roman 
formulas are of the type erovs r\ avTOKparopos Kata-apoi Tltov 
AIXlov ^Abpiavov 'AvTcavivov Evcrej3ovs Se^ao-roS, Me)(etp e, 
the shorter eroi'? ij ^ Avtcovu'ov Kaicrapos rov Kvpiov. In 
neither case can there be any doubt as to the precise 
date intended, and it is only under some of the later 
emperors that the accumulation of titles {Tepp-aviKov 
YlapOiKov Aa/ci/coC ^peravviKov Meyto-roi», and the like) 
becomes cumbersome. 

This useful and practical system of dating was, however, 
abandoned at the time of the revolution under Diocletian. 
In the first instance dating by the consuls of the year 
was substituted; but in the year 312 the system of the 
indiction was instituted. This was a fifteen-year period, 
beginning on different days in different parts of the 
empire. At Constantinople it began on the ist September, 
in Egypt on a fluctuating date about the middle of June, 
at the time of the rising of the Nile ^ Its origin is 
uncertain, but it is probably a modification of the fourteen- 
year census-period in use under the Romans. The formula 
of dating by the indiction is ^appovOi (, t? irStKnwro?, 
meaning the fifteenth year of the current indiction-period : 
but as the indiction-periods are not themselves numbered, 

' See Greek Pajnjri in f/ic British Museum, i. 196-8. 


this dating is absolutely useless from our point of view. 
Sometimes the names of the consuls or the regnal year 
of the emperor are added, but this is not by any means 
the regular practice ; and consequently the dating of 
Byzantine documents remains an obscure subject, in spite 
of the immense mass of material for the period which is 
in existence. The publication of select dated facsimiles 
from the Vienna and Berlin collections is much to be 
desired, in order to remove this obscurity ; but perhaps 
a speedier revelation may be looked for from the Egypt 
Exploration Fund. 



In passing to the consideration of literary papyri, one 
enters what is at once the most important and the least 
familiar part of the subject. Non-literary papyri, as will 
have been seen from the preceding chapter, are plentiful 
and fairly well supplied with precise dates ; literary papyri 
are comparatively rare and can seldom be dated, even 
approximately, on other than purely palaeographical evi- 
dence. Nor can all manuscripts which contain literary 
v/orks come properly into consideration here, since some 
of them are not written in formal book-hand at all, but 
in the ordinary private or non-literary hand of the day. 
Such is, notably, the papyrus containing Aristotle's Consti- 
tution of Athens; also a large medical manuscript in the 
British Museum, containing considerable extracts from the 
'larpiKi] ^vvayojyi] of Menon, the pupil of Aristotle, some- 
times ascribed to Aristotle himself; and an astronomical 
treatise derived from Eudoxus in the Louvre, of the second 
century B. c. In this way the bulk of the materials at our 
disposal is reduced ; but on the other hand there are a few 
documents of a private or business character which are 
written by professional scribes in literary hands ; and these, 
bearing, as they do, precise dates, are very valuable items 
of evidence for the construction of the history of papyrus- 
palaeography on its literary side. 

Indeed the document to which is generally assigned 


the first place in any treatment of the subject is itself, 
to some extent, one of this description, being a non-literary 
composition written in an uncial hand. This is the so-called The 
'Curse of Artemisia,' consisting of the imprecations of Artemisia. 
a woman against the father of her child, who has de- 
serted her, written in an extremely archaic uncial hand. 
The papyrus was found on the site of the Serapeum 
at Memphis (though not with the papers of the recluse 
Ptolemy, mentioned in the last chapter), and is now in the 
Imperial Library at Vienna ^ It has been assigned by 
Blass and Wessely to the fourth century e. c, by Thompson 
to the early part of the third ; and there is no doubt that 
the forms of the letters, recalling as they do those of 
inscriptions upon stone, have a veiy early appearance. 
The letters are stiff and angular, with few curves; A 
and E are of the square epigraphic shape, the latter 
generally having the top horizontal stroke very long; 
© is a circle enclosing a dot ( O ) ; O is small ; C is of a form 
intermediate between the Z of the inscriptions and the C of 
the papyri (C, (); OJ retains much of the epigraphic form (^ , 
v/t) ; a colon is used for purposes of punctuation, as in in- 
scriptions and a few early papyri. What militates against 
the value of this document for palaeographical purposes 
is its extremely rough and untrained appearance. It is 
not the work of a professional scribe, but the writing 
of an uneducated woman who uses uncial letters because 
she can form no others. If Greek writing were at this 
period just issuing from the purely epigraphic stage, stress 
might fairly be laid upon the use of epigraphic forms 
as a proof of age ; but since we now know that men wrote 
freely and easily upon papyrus long before this time, this 
argument falls to the ground. Artemisia used letters like 
those employed in inscriptions for the same reason that 
an illiterate person always uses capitals, because such 

' Facsimile in the Palaeographical Society's publications, vol. ii. pi. 


letters were commonly before her eyes in public places, 
while she had probably seldom seen a book. It is on 
evidence of a different character that the early date of this 
document can alone be maintained. The traces of Ionic 
dialect ('Apre/ztfrtrj, tKerrjpi?]}, the assimilation of the final 
consonants of prepositions to the first letter of the following 
word (fju TToa-epdiTi, ey yf/t), and the occurrence of forms 
of letters which are found not only in inscriptions but in 
papyri known to belong to the third century, such as those 
of €, 0, Ol), described above, may be admitted as legitimate 
evidence that this is a genuinely early document ; but it is 
not one on which we could base any sound argument 
as to the character of contemporary^ MSS., if we had no 
knowledge on this point from other sources of information. 
The third Our knowledge of the literary palaeography of the third 
B^c."^^ century b. c. is based, in fact, entirely on the papyri 
The Petrie discovered by Mr. Flinders Petrie in the mummy-cases 
of Gurob, to which reference has already been made more 
than once. Among them, in addition to the non-literary 
documents described in the last chapter, were several 
fragments of literary works ; and it is a fair conclusion 
that these are of about the same age as the non-literary 
papers among which they were found. The argument 
is, of course, not quite decisive. The papyri out of which 
the Gurob cartonnages were made were, no doubt, waste 
paper at the time when they were so used ; and it may be 
questioned whether literary and non-literary documents 
found in the same waste-paper basket are likely to be 
of the same age. On the one hand it has been argued 
that a well-written MS. of Plato or Euripides would 
not be thrown away nearly so soon as mere business 
papers of ephemeral interest, and hence that the fragments 
here in question may easily be as early as the fourth 
century. Against this it has been contended that such 
documents as wills, leases, and receipts, which form the 
titles to property, would probably be preserved for several 


generations, while a copy of a literary work mio;lit he 
thrown away at any time if it M^ere found defective 
or injured in any way. These arguments may perhaps 
be allowed to neutralize one another. Titles to property 
do not seem to have been preserved with the same tenacity 
as in modern days, and were indeed unnecessary when 
a register of ownership was kept by the government ; 
while, if some of the literary MSS. M^ere of a period much 
earlier than that at which they were converted into 
mummy-cases, we should expect to find a perceptible 
gradation of hands in the various fragments, some being 
old while others were more recent. This is not the case 
to any important extent; and we are therefore justified 
in assigning the literary fragments among the Petrie 
papyri to the same period as their non-literary companions, 
namely to the third century. 

To the Petrie fragments may now be added a few scraps 
acquired by Mr. Grenfell in 1895-6, which, though bearing 
no independent proof of their date, evidently belong to the 
same period ^ What we have, therefore, in all, as our 
material for judging of the literary palaeography of the 
third century, consists of the following : four columns of 
the A ntiope of Euripides '^ ; about twelve columns, besides 
minor fragments, of the Phaedo of Plato ^; five columns 
of the Laches of Plato * ; and about a score of minute 
fragments of various works'", all of them unknown with 
the exception of a few pieces of Homer. Some of these, 
however, can hardly be reckoned as literary hands. With 
these, and with the evidence derivable from the non-literary 

' Facs. in Grenfell, Greek Papyri , vol. ii. pi. i ; now in British Museum. 

^ Facs. in Mahaffy, Flittders Petrie Papyri, part I. pll. i, ii ; now in British 

^ Facs. ih., pll. v-viii ; partial facs. in Plate X and in Pal. Soe. vol. ii. 161 ; 
now in British Museum. 

* Facs., Mahaffy, part II. pll. xvii, xviii; now in Bodleian Library. 

* Facs., ib., part I. pll. iii, iv, ix, x, xxv ; part II. pi. xvi ; and Grenfell, 
1. c. ; now divided between the British Museum and the Bodleian. 


papyri of the same period, the history of writing during 
the century has to be reconstructed. 

An examination of these texts shows, in the first place, 
that they fall into two groups, one consisting of very 
small hands, while the others are larger and present 
generally a rougher appearance, apart from differences 
in the formation of individual letters. The more impor- 
tant, and also the best-written, texts are all in the smaller 
hand, which may fairly be taken as representing the 
writing in current use for books intended for public 
circulation. They include the Phaedo and Antiope MSS., 
a few of the Petrie fragments, and all those of Mr. Grenfell. 
The oi-der of antiquity among them can only be deter- 
mined by the more or less archaic appearance of the 
writing. In the beginnings of writing on papyrus it 
cannot be doubted that, in formally and carefully written 
MSS., the shapes of the letters were nearly identical 
with those in contemporaneous use in inscriptions; and 
the greater or less occurrence of epigraphic forms, in 
a MS. written, not by an uneducated person (as in the 
case of the Artemisia papyrus), but by a trained scribe, 
may be taken as evidence for a relatively earlier or 
Early later date. Applying this test, the earliest examples 
fragments. ^j^-^Q^g the papyri now extant would seem to be (as 
Mr. MahafFy and Mr. Grenfell have already noticed) a frag- 
ment of a prose narrative of the adventures of Heracles, 
among the Petrie papyri \ and two minute scraps among 
Mr. Grenfell's. The most notably early among the forms 
of letters in these fragments are the square E with broad 
top ; with a central dot instead of a cross-stroke ; (x) in its 
transition forms, between H and co ; and especially Z, 
formed of three parallel horizontal bars with a perpen- 

 An ali^liabet from this fragment is given by Mahaffy in his Flinders 
Petrie Papyri, part I, p. 65 ; and an alphabet from the lai-gest of Mr. Grenfell's 
fragments foi-ms col. i of the table given in Appendix I to this essay, 
to which the reader may be referred for illustrations of the descriptions 
of the individual letters which follow. 


diciilar stroke cutting them at right angles, and C in the 
angular epigrapliic shape. These last two characters belong 
only to the Grent'ell fragments ; the Heracles fragment has 
no example of a Z, and its C is semi-circular. Of the other 
letters it is sufficient to observe that A is of the uncial 
shape, like a modern capital A ; H is rather rounded 
on both sides ; the two strokes composing A meet at the 
top ; M is somewhat deeply curved in the Grenfell fragment 
and angular in the Heracles ; O is very small ; V\ is rather 
broad, with the right leg inclined to be shorter than 
the other ; and Y is shaped as here printed. Not all these 
letters are of value for determining the status of these 
texts among other third-century, or even among Ptolemaic, 
papyri ; but it is necessary to note them for comparison 
with the forms which came into existence in later periods. 

Somewhat later than these fragments — perhaps about The 
the middle of the third century — must be placed the two an^j ^„. 
great treasures of Mr. Petrie's find, the Phaedo and the ''opeMSS. 
Antiope. In these cases we have some substantial part 
of each MS. preserved, and are in a better position to 
judge of their general appearance. The hand is not the 
same in both, that of the Phaedo being noticeably/ better 
and more careful ; but in size and formation of letters 
they are alike. The writing is extremely small, with 
the object, no doubt, of making neat and handy volumes ; 
the complete Antiope, supposing it to have been a play 
of between 1400 and 1500 lines, would have occupied a roll 
of about 12 feet in length. Of the individual letters^, 
A is of the uncial shape, in the Antiope sometimes (but not 
always) approaching nearer to the epigraphic shape by 
having the cross-bar bent downwards into an angle. The 
Plato also sometimes has the square E with elongated top ; 
but elsewhere in the Plato, and always in the Euripides, 
the rounded form of the letter is found. Z (which does 

' See alphabets in columns 2 and 3 of the table in Appendix I. 


not occur in the earlier fragments described in the last 
paragraph) has the form, peculiar to the third century, 
of two parallel bars joined by a perpendicular stroke ; 
has a cross-stroke, though it does not always accurately 
fit the circle ; A and M are of the forms already described ; 
Z is composed of three parallel and unconnected strokes, 
and this, it will be found, is characteristic of the Ptolemaic 
age alone, and forms a useful criterion for distinguishing 
Ptolemaic MSS. Of the remaining letters, fl is very 
broad in the Phaedo (it will be remembered that this 
is a feature of third-century non-literary papyri), less 
so in the Autiope; C is small and rounded, the top some- 
times slightly flattened ; (jl) is of the minuscule shape, the 
second loop being often incompletely formed. 

The beauty of these hands, and especially of the Plato, 
has been eulogized in terms which are, perhaps, somewhat 
exaggerated ; not unnaturally, since, in the state of our 
knowledge at the time of their discovery, it was a surprise 
to find writing of so early a date which showed such 
freedom of style, combined with so much orderliness 
and precision. Both Mr. Mahafi'y and Sir E. Thompson 
speak of the Plato, in particular, as a MS. of extreme 
beauty. Now that the first joy of discovery has worn off, 
it seems evident that they are not the equals of the best 
specimens of the Roman period, or even, it may be, of the 
later Ptolemaic age. There is a general appearance of 
neatness about these small, yet firm, hands; but there 
are considerable inequalities in size among the letters, 
which would be more noticeable if they were written 
on a larger scale, and they lack the handsomeness of the 
larger hands which came into favour later. The Amt'iope 
MS., indeed, hardly achieves the praise of neatness, being 
decidedly rough and irregular in places, besides being more 
compressed than the Plato. It cannot be supposed that 
this is a fair example of the volumes preserved in the 
great Alexandrian library, though it may represent the 

Plate X. 

4*^ h-rr^H^ 

J" -«\ 1 *^ ^A 

'-fr^f --°f f «-&-•*'•* ^•'^i^'i vests' ^■^"■'^ 




-'vr- ►^■o ^t^M'r^}* :^;^J?^AJtfi>-i■*7^^^a 

' ->■' 

• ''■^""'■'1 ^ '" -'- " U^-*'- ' ■"';.. .,;-■;---' ./-"'-■". 


.:^ ';' 

^ , 





style which would pass muster in the book-shops. Nor 
do the texts come up to the Alexandrian standard of 
accuracy. How far they fall short is a question to which' 
the answer depends on the solution of the critical problem 
as to the authenticity of certain readings which depart 1 
from the vulgate ; but apart from these there are more ' 
obvious blunders than can have been admitted in the great 
libraries which handed down, in most cases, such pure 
texts to the vellum MSS. of the tenth and later cen- 
turies, on which our present knowledge of the classical 
authors is based. Yet, with all these deductions, the " 
Phaedo and Antiope papyri have a great attractiveness 
of appearance, and rank high among papyrus MSS. 
of any age ; nor would any one deny their extreme value 
for the history of Greek writing. With their aid we can 
not only realize how Greek books were written in the days 
of the early Ptolemies, but can, by a legitimate use of 
inference and imagination, picture to ourselves the contents 
of the book-shops of Athens in the times of Menander and 
Demosthenes, perhaps even of Aristophanes and Sophocles. 
Between the literary hands which have hitherto been 
described, and the non-literary hands of the same period, 
there is no very marked resemblance. Some of the letters, 
such as n and Z, and to a less extent M, T, and Y, would 
be recognized as Ptolemaic in style, but it would be 
very difficult to assign these MSS., on palaeographical 
grounds alone, to the same perio<l as the non-literary 
documents which bear dates in the reigns of the early 
Ptolemies. There is, however, another type of hand to 
which this statement applies much less fully. Several 
of the literary MSS. among the Petrie papyri are 
in larger hands, approximating to those of the con- 
temporary non-literary documents. Chief among these The 
are the fragments of the Laches of Plato, now at Oxford, jj'^'' !,n(i 
the curious scrap of Homer containing several additional others, 
verses, the fragments of the Mouseion of Alcidamas and 


another rhetorical treatise ^. The other classical fragments 
published by Prof. Mahaffy likewise belong to this category, 
but they are so small, and in some cases approach so 
closely to the non-literary hand, that they can hardly 
be taken into account. In this class of hands, while 
the uncial forms of some letters, such as A, H, K, N, are 
preserved, and the straggling appearance of the non-literary 
hands is avoided, the writing on the whole is recognizably 
akin to the non-literary type. The letters are larger 
and squarer than in the Phaedo and Ant lope; C, Oi), and 
in some cases M, are definitely of the minuscule form; 
and the writing is without the style and attractiveness 
— due, no doubt, to greater care — which characterize the 
smaller hands described above. The MSS. of this larger 
type cannot be regarded as having been intended for 
sale or general circulation. They are rather the work 
of ordinary scribes, employing a somewhat modified 
form of the current handwriting of the day, which they 
have adapted for literary purposes by reducing its excessive 
breadth and cursiveness, making it squarer and firmer, 
without, however, achieving much of beauty or regularity. 

The For the second century b. c. materials are, at present, 

century very scarce. There has been no great discovery of 
^- ^- documents of this period, comparable to the Petrie ' find ' ; 

and the principal groups of non-literary material which 
have come to light — the Serapeum papyri for the first 
half of the century and Messrs. Hogarth and Grenfell's 
papyri for the second half — contribute between them only 
one MS. which can in any way be classed as literary. In 
addition to this single MS. there is also one isolated 
discovery belonging to this period, which is of considerable 
interest. Both MSS. are now in the Louvre, the last- 
mentioned, a recent acquisition, being the famous oration 
of Hyperides against Athenogenes — one of the orator's 

' Petrie Tapijri, part I, nos. lo, 25. 


masterpieces — while the other, ac(|iiired at a much earHer 
date, is a dialectical treatise containing quotations from 
a number of classical authors, such as Sappho, Alcman, 
Ibycus, Homer, and Euripides. Both MSS. are shown 
to be not later than the second century B.C. by the 
existence of writing on their backs belonging to the 
Ptolemaic period : while neither of them shows any such 
marked resemblance to the Petrie papyri as would justify 
the assignment of them to the third century. The 
dialectical treatise must, however, belong to the first 
half of the second century at latest, since the writing 
on its veri<o is dated in the year 160 B.C. For the 
Hyperides the only evidence we have to go upon is the 
statement of Prof. E. Revillout, its first editor, that it 
bears on its verso accounts in demotic, belonging to the 
Ptolemaic period. No more precise indication of their 
date has been given ; but this is sufficient to show that 
the text of the oration can hardly be later than the 
second century, while certain resemblances between its 
writing and that of MSS. belonging to the following 
century seem to make it probable that it belongs to the 
second half of the century rather than the first. From 
only two MSS., and these exhibiting very different types 
of writing, it is obviously dangerous to draw any very 
sweeping conclusions as to the literary palaeography of the 
period ; but they serve to bridge over the gulf between 
two periods for which our information is more complete. 

The dialectical fragment ^ is written in narrow columns The Paris 
of about two inches in breadth, leaning (as in several Fragment, 
other early papyri) strongly to the right. Divisions in 
the sense are marked by slight spaces in the writing, 
and by |x( 7^1:^7 'cyjAi below the beginnings of the lines in 
which the pauses occur. The hand is small, though not 

' Complete facsimile (not photographic) in the Album to Notices et 
Extraits, plate xi ; photographic facsimile of three columns in Pal. Soc. ii. 
180, and of one column in Plate XI. Alphabet in Appendix I, col. 4. 



so minute as those of the Phaedo and Ant lope ; it is 
evidently to be regarded as the lineal successor of these 
hands. The most noticeable modification in general form 
is a tendency to roundness and curved forms of letters. 
In this respect a transitional form may be found in a 
small fragment discovered and published by Mr. Grenfell ^ 
The letters are also more even and uniform in size ; and 
in many of their forms they approximate to some of 
the MSS. found at Herculaneum, which must be at 
least a century later. Of the individual letters, A is the 
most noticeable, as giving the first example of a form 
found in some other early papyri. It is a modification 
of the uncial form, the original left-hand stroke and the 
cross-bar being formed together into the shape of a sort 
of wedge {^), across the open end of which the right-hand 
stroke is drawn, generally in a curve (a). It is an inter- 
mediate form between the uncial (A) and the fully 
developed minuscule (d^), which can be written without 
raising the pen. Another letter which shows a transitional 
form is Z, in which the middle stroke is neither perpendicular 
as in the Petrie MSS., nor drawn obliquely from end to 
end of the horizontal strokes as in later papyri, but is 
intermediate Ijetween these two forms. H and K are 
strongly curved. A is noticeable for its right-hand stroke 
in some cases (but not in all) being carried beyond the 
point of junction ; again a sign of transition, since this 
is generally characteristic of late Ptolemaic and Roman 
MSS. Z is still decidedly Ptolemaic, being formed of three 
distinct strokes, the middle one of which is very small ; 
so also is n, which generally retains its breadth. C is 
well rounded ; and the curve of Y is more strongly marked 
than before. The other letters have no peculiarities that 
need be noticed. The general aspect of the writing is 
neat and graceful ; less strong than the best third-century 
hands, but somewhat more ornamental. 

' Greek Papyri, ii. pi. i, no. 2 ; now Pap. DCLxxxixa in the British Museum. 

Plate XI. 




/ f 

* >j<J^fer^Woyjco7-4^-y>n^ 

^^^cJcA/ofrr^a^Xf JSfOJUc >fKoyKoV^on" ^^OdNfTTf^ 


Plate XII. 

"='33:-, . *- ^"ST*- >.- •* • >«'. , 

■* .- >- 



The Louvre Hyperides ^ presents a very different The 
appearance from its colleague. If the main characteristics Hyper- 
of the latter are grace and roundness, those of the ''^^'*- 
Hyperides are strength and squareness. The columns 
are broad and upright, instead of being narrow and 
sloping. The letters are firm and noticeably square in 
appearance, besides being larger than those of the dialectical 
treatise. The uncial forms of such letters as A, H, M, are 
retained, and all the characters are fully and carefully 
formed. This MS. is, in fact, the first example of 
a phenomenon which runs all through the history of 
literary papyri, namely the retention of a fully developed 
uncial form of writing side by side with hands in which 
a certain concession to the cursive style is discernible. 
The dialectical treatise is an example of this intermixture 
of the cursive element, while the Hyperides rejects it. 
It is obviously more difficult to assign dates to the strict 
uncial hands, which approach the epigraphic style rather 
than that of non-literary documents ; but there are some 
letters on which dependence may generally be placed. 
In the present instance, Ptolemaic characteristics are 
unmistakably to be found in A, the right-hand stroke 
of which projects very slightly, or not at all, above the 
point of junction; in M, which has a shallow, angular 
depression ; in the three unconnected strokes of Z ; and 
(less distinctively) in the Y-shaped form of Y. Of the 
other letters it need only be noted that A and A have 
the peculiarity of a short projecting stroke above the 
apex ; that Z sometimes has its down- stroke drawn to 
the middle of the lower bar, instead of to its left-hand 
end ; and that € has the cross-bar rather high, while the 
upper curve is carried well over, so as nearly or quite 

' Complete facsimile with printed text, edited by E. Revillout (1893) ; 
one column in Hyperides, the Orations against Athenoqcnes and Philippides, 
edited by F. G. Kenyon ; another in Plate XII, Alphabet in 
Appendix I, col. 5. 

F 2 


to meet it again. The general appearance of the hand 
is attractive and imposing, and, although less economical 
of space than the minute hands of the Phaedo and Antiope, 
it has good claims to be considered a more handsome style 
of writing. It is, moreover, very distinct, and therefore a 
valuable medium for the safe transmission of literary texts. 
The Passing from this, the finest example of second-century 

mad. calligraphy, there are two other literary papyri of the 
same period which deserve a brief mention. The first 
of these is a fragment containing a small portion of the 
eleventh book of the Iliad, in which, as in the Petrie 
fragment mentioned above, there are some additional 
verses which are not found in the vulgate. The papyrus 
is at Geneva, and the text has been published by Prof. 
Nicole, who, however, does not assign any date to it. 
From a fascimile, however, published by Prof. Diels ^, 
it is clear that it belongs to the second century B.C., 
being written, in fact, not in a literary hand at all, but 
in a reduced form of the common non-literary hand 
which meets us in the Serapeum papyri ^. The question 
as to the character of the enlarged text contained in it 
does not come within the scope of the present essay ; 
though it may be observed that this is the latest example 
of such a text which has yet come to light. Some 
additional specimens of it have lately been acquired by 
Mr. Grenfell, but they, like the Petrie fragment, belong- 
to the thiixi century B.C. 
The Erotic The other literary papyrus alluded to above is the 
ragment. fragment of a mime, written either in rhythmical prose 
or in lyrical verse without strophic correspondences (for 
scholars differ), recently published by Mr. Grenfell ''. This 

' Sitzungsherichie der k. prevssischen Akademie, 1894, no. xix. 

- It is the more necessary to record the date of this MS., because Pruf. 
Maliaffj", in a moment of aberration, has assigned it to the second century 
after Christ. 

' Greek Fajjyri, vol. i. {An Alexandrian Erotic Fragment), no. i. Now Brit. 
Mus. Pap. Dcv verso. 


unquestionably belongs to the second century B.C., but 
is written in a non-literary hand, and therefore does not 
require separate treatment here. It is inscribed on tlie 
verso of a document dated 173 B.C., in a small cursive 

There are other papyri which have sometimes been con- 
jecturally assigned to the second century B.C., notably 
two of the British Museum MSS. of Hyperides ; but in 
the pages which follow these MSS. have been referred to 
other periods, and will not be discussed until we arrive 
at the dates to which, according to the evidence now 
available, they appear to belong. 

From the scantiness of evidence which is thus seen 
to be available, it is impossible to formularize any very 
dogmatic conclusions as to the literary palaeography of 
the second century b. c. ; especially as the two MSS. which 
form the principal sources of our knowledge exhibit very 
different types of writing. It is fortunate, however, 
that the approximate dates of these two MSS. are estab- 
lished with more certainty than usual, by the writings 
which each has upon its verso. We are thus in possession 
of fixed points from which other examples may be dealt 
with ; and in approaching the doubtful period which follows 
next such fixed points are very valuable. 

The first century B.C. is, indeed, a period which requires The first 
careful handling and little dogmatism, both because several ^^^^^^''^ 
interesting MSS. have been assigned to it with various 
degrees of doubtfulness, and because it contains the 
transition from the Ptolemaic to the Roman periods. 
A special element of complexity is introduced by the 
Herculaneum papyri, which, from their character and 
the circumstances of their preservation, must necessarily 
belong either to this century or the next, and which, for 
the first and last time, carry the range of available 
evidence beyond the borders of Egypt. There are a few 
fixed points which mark out the stages of the inquiry, 


and round these the attempt must be made to group 
the other documents which, on palaeographical grounds 
alone, appear to belong to the period. 
The Her- The Herculaneum papyri, with which it will be con- 
Papvi-i"'" venient to deal first, raise a question which has not 
yet been discussed, namely how far the history of palaeo- 
graphy in Egypt applies to the Greek-writing world 
outside. The question cannot be fully answered, for 
want of sufficient materials ; but it can be answered 
sufficiently for all practical purposes. The analogy of 
mediaeval Latin MSS. shows that the distinctions be- 
tween contemporary MSS. written in difierent countries 
are less than those between MSS. written in the same 
country but at the distance of a century or so of time. 
It requires more experience to determine the country of 
a MS. than its date. There is no reason to doubt 
that the same would be the case with Greek writing on 
papyrus ; on the contrary, there is the more reason to 
expect uniformity, since all Greek MSS., in whatever 
country, would be written either by Greeks or by those 
who had learnt their writing from Greeks. We have, 
in fact, practically the same conditions as prevailed in 
the Greek minuscule writing of the late Middle Ages. It 
may be possible for special experience sometimes to dis- 
tinguish Greek MSS. written in Italy from those written 
in Greece and Constantinople, but it is not necessary to 
possess this special knowledge in order to assign the date 
of a MS. with sufficient accuracy. 

When, therefore, we find that the papyri discovered 
at Herculaneum, though not exactly like any of the 
Egyptian papyri, yet do not differ from them more than 
they differ among themselves, and possess strong re- 
semblances both in general appearance and in particular 
detail, there need be no hesitation in applying to them 
the same criteria of age as to their Egyptian relatives. 
No doubt there may have been local distinctions, which 


fuller evidence might enable us to discern ; and certain 
features may have continued longer or less long in 
Alexandria or in Rome ; but this range of uncertainty 
is no greater than that which applies to the whole 
subject of supplying conjectural dates to undated MSS., 
and in practice may be ignored. It may, however, 
be worth while to bear in mind that Greek writingf was 
an exotic among the Romans, while it was naturalized 
in Egypt ; hence it is likely that changes in the prevalent 
styles of writing began rather in Alexandria than in 
Rome, so that j)ossihly MSS, written in Italy may be 
slightly later in date than MSS. showing the same 
characteristics written in Egypt. Since, however, the copy- 
ists of Greek manuscripts written in Italy were almost 
certainly Greeks themselves, not much weight need be 
attached to this argument. 

It seems, therefore, perfectly legitimate to use the Their 
Herculaneum papyri as evidence for the palaeography '^ ^' 
of MShS. written in Egypt ; and the point is of consider- 
able importance, since the date of the Herculaneum 
volumes can be fixed with fair accuracy. The terminus 
ante quern is absolute, being the eruption of Vesuvius 
in A.D. 79, which overwhelmed the town and calcined 
the papyri. The terminus a quo is supplied by the 
fact that many of the papyri contain works of the 
Epicurean philosopher, Philodemus, who was a contem- 
porary of Cicero ; while nearly all the rest are either 
copies of the works of Epicurus or treatises of a philo- 
sophical character. Now Philodemus was not a very 
eminent philosopher, and it is not likely that a large 
collection of his works would be made at any time much 
after his death ; and the fact that several of his treatises 
are here represented in duplicate suggests the probability 
that the collection was made by Philodemus himself ^ This 

' See Scott, Fragmenia Herculanensia, pp. 11, 12. 


would assign the MSS. to the first century B.C., and 
to dates before, rather than after, B.C. 50. This con- 
clusion is not invalidated by the fact that a Latin 
poem on the battle of Actium and the Egyptian campaign 
of Augustus was found along with the Greek papyri, since it 
may easilj'' have been added by a later owner of the library. 
On the other hand, this poem is itself not at all likely to 
have been written later than the reign of Augustus, and 
therefore supplies a proof that the library was not of 
recent formation at the time of the catastrophe of a.d. 79. 
A philosophical library of the preceding century may 
very well have been kept together in the house of its 
original owner, just as libraries of eighteenth-century 
theology may be found in country houses nowadays, 
even though some of the authors are somewhat out of 
date. There is, therefore, no ground for rejecting the 
natural inference from the character of the library itself, 
to the effect that the volumes composing it were written 
shortly liefore the middle of the first century b. c. 

The Herculaneum papyri consequently belong to the 
end of the Ptolemaic period of Greek palaeography ; for 
it will be convenient to retain the name, although it is 
something of a misnomer when applied to MSS. written 
outside Egypt. Indeed, as has been briefly indicated in 
an earlier chapter, the names ' Ptolemaic,' ' Roman,' 
' Byzantine,' must be applied much less rigidly to literary 
than to non-literary papyri. Changes of dynasty, im- 
plying changes of influence and the introduction of clerks 
and officials of different origins, may very naturally affect 
the handwriting of official, and through them of private, 
documents to a very noticeable extent ; but the writers 
of books belong to a society which extends beyond the 
borders of a single kingdom, and would not consciously 
submit to the influence of an official chancellery. Hence, 
on the one hand, the changes in literary papyri at the 
advent of the Romans in Egypt are less obviously recog- 


nizable ; on the other hand, they may be taken as applying 
generally to Athens and Rome as well as to Alexandria. 
In speaking of any Egyptian papyrus, written in a formal 
book -hand, as ' Ptolemaic,' we mean only that it is written 
in a hand prevalent in Greek literary circles during the 
period when the Ptolemies ruled in Egypt, with, perhaps, 
some sliofht local modifications due to the fact that the 
scribe was writing in Egypt ; and similarly, in sj^eaking 
of papyri written at Herculaneum as Ptolemaic, we mean 
that they are written in the Greek book-hand prevalent 
during this same period, even though the slight local 
inlluences were in this case Italian and not Egyptian. 

The conclusion that the Herculaneum papyri are to be Test 
referred to the end of the Ptolemaic period suits perfectly |f^|^f^|f': 
with what we know from other sources of the course of 
palaeographical development, and they thus drop easily 
into a natural place in the sequence. In all crucial details 
they answer to the criteria which serve to distinguish 
Ptolemaic from Roman papyri ; and the history of Greek 
palaeography remains natural and intelligible on this theory, 
which it is not if they are referred to a date a century 
or more later. The two letters which are of most decisive 
importance at this stage are A and Z. The uncial form of 
the former letter (A) does, indeed, as has been stated above, 
run through all periods, and is found alike in Ptolemaic 
and Roman MSS. ; but the minuscule form is different 
in the two periods. In literary papyri of the Ptolemaic 
age the right-hand oblique stroke is always formed sepa- 
rately from the rest of the letter, which is written without 
lifting the pen, having either an acute angle or a loop in 
the left-hand corner (A., ^v, cA., A). In Roman papyri, 
on the other hand, the A is almost always written in one 
piece, just as in cursive MSS. (J-, ^). There are exceptions, 
as in the case of the Bankes Homer (Brit. Mus. Pap. cxiv) ; 
but they are few, and the rule will generally be found 
serviceable, except in the cases where the uncial form of 


the letter is retained. The second test letter is still more 
valuable, being liable to no such exception as that just 
mentioned. It is, of course, rash to rely absolutely on the 
form of a single letter as an infallible test of date, especially 
where evidence is so scanty as it is in the case of literary 
papyri ; but so far as the evidence goes at present, Z in 
Ptolemaic MSS. is invariably formed of three disconnected 
strokes \ while in Roman MSS. it is equally invariably 
formed in one continuous whole. This rule at present holds 
good absolutely, though in paleograjohical matters isolated 
exceptions to any rule may always come to light. 

According to both these tests, the Herculaneum papyri 
fall within the Ptolemaic period. They include, as might 
be expected, several ditlerent hands, but most of them 
exhibit the same general type - ; and in respect of these 
special letters there is little variation. A is normally of 
the minuscule type, formed in two strokes ; in a few cases, 
such as the MS. of Philodemus De Ira, it is of the uncial 
shape ; but never, apparently, of the Roman type, written 
without lifting the pen at all. The evidence of Z is equally 
clear and unanimous, the letter being regularly formed 
of three distinct strokes. The only variation is that in 
some cases the central stroke is a horizontal line (shorter 
than the two others), while elsewhere it has the shape 
of a comma (Z, X). In one MS.^, if the facsimile (which 
is not good) is to be trusted, the central stroke does 
indeed touch and connect the two others (x) ; but the 
exception is more apparent than real, since the letter 

' One of Mr. Grenfell's fragments, described above (p. 45) has the three 
strokes connected by a perpendicular line at right angles to them (I) ; but 
this is a still more archaic form, and does not in any way invalidate the 
principle here enunciated. 

2 See facsimiles in Scott's Fragmenta Herculanensia ; Thirty-six Engravings of 
Textsand Alphabets from the Herculaneum Fraginents{<dA..W\ch.o\son,OyiiovdiQgi) ; 
Pal. Soc. i. 151, 152; and a series of photographs issued by the Oxford 
Philological Society, in several volumes. Alphabet in App. I, col. 7, below. 

^ Scott's Pap. 26. 


is still formed by three distinct strokes of the pen, the 
only difference being that the central comma-shaped stroke 
is large enough to touch both the upper and the lower 
horizontal line. 

In general appearance the Herculaneum hands are 
small and graceful, having affinities with the Louvre 
dialectical fragment described above, rather than with the 
larger and squarer type of the Louvre Hyperides. A, as 
will be seen from the above description, is not unlike the 
same letter in the dialectical papyrus ; £ is generally well 
rounded, and the central stroke has a tendency to be 
separated from the rest of the letter; the two strokes 
of A are generally united at the top, but sometimes the 
right-hand stroke is prolonged ; M is Ptolemaic, having 
an angular centre instead of the deep rounded curves of 
the Roman period ; fl is fairly broad ; Y is of the Y-shape, 
but without any strong individuality. All these details 
tend to confirm the assignment of these papyri to the 
end of the Ptolemaic period, and assist in giving fullness 
ar^d precision to our knowledge of its palaeographical 

The Herculaneum papyri having thus been assigned Tlie 

to their place, two other MSS. must be mentioned which Bacchy- 

likewise seem to belong to the earlier half of the first Papyrus. 

century B. c. The first of these is the recently discovered 

MS. containing the poems of Bacchyhdes^. This has 

strong Ptolemaic features; but something in its general 

appearance, as well as some details which it has in 

common with later MSS., indicates that it probably 

belongs to the end of the period, about the middle of the 

first century b. c. It is in a handsome hand, of good size, 

clear and firm, and carefully written. It evidently belongs 

to the class of MSS. written for sale or for preservation 

^ Complete facsimile published by the Trustees of the British Museum. 
One column is reproduced in Plate XIII, and an alphabet given in App. I, 
col. 6. 


in a public library, and ranks high among the extant 
specimens of Greek calligraphy upon papyrus. Its columns 
are wide (about 5 inches), their size being determined by 
the length of the longest of the irregular verses in which 
the odes are written. Of the individual letters, A is con- 
spicuously angular, without a vestige of a curve, and 
formed in two strokes of the pen, (A, K)\ £ is narrow, 
wdth a long projecting central stroke ; Z has the peculiarity 
of often having the lower horizontal stroke separate from 
the rest of the letter; A shows a slight projection of the 
right-hand stroke, which may be taken as an indication 
of the approach of the Roman period ; M, on the other hand, 
is decidedly Ptolemaic, being a broad letter with a shallow 
curve; Z, too, is quite Ptolemaic, and the central bar is 
generally reduced to a mere point, while the other strokes 
are unusually long ; O is small ; PI is broad ; C is narrow, 
and the upper part of it is often separated from the rest, 
a phenomenon which recurs in the Harris Homer, to be 
described in the next chapter ; T has its cross-stroke longer 
on the left than on the right ; Y has a very shallow curve ; 
(jj is broad and shallow, with little or no traces of the 
central stroke. Some of the titles of poems have been 
added in a different hand, which is plainly of Roman 
type, and probably not earlier than the second century ; 
but these are so clearly a later addition, being omitted in 
some instances, and showing quite distinct forms of letters, 
that they give no clue to the date of the original writing \ 

^ The (late here assigned to the Baechylides papyrus is questioned 
by Messrs. CTienfell and Hunt .Oxyrhijnchus Papyri, i. 53), who would place 
it in the first or second century after Christ, partly on the strength 
of a papyrus of Demosthenes published by them, which (having accounts 
on the rerso in a second-century hand) they assign to the early part of the 
second century. The resemblance between that papyrus and the Baechy- 
lides, however, is not very great, and to my eye the Demosthenes appears 
decidedly later. None of the most characteristic letters of the Baechylides 
(fx, £, V, cu ) has the same shape in the Demosthenes. Personally I should 
put the Demosthenes fragment somewhat earlier than they do, regarding 
the cursive writing on the verso as of the middle of the second century, 

Plate XI 1 1. 

1 / 

/ B-ff f^^'^/^^>c>>.^J-fT^M^-f^^r^M^^i- 
t-TMM tT^KHVTvHirff Hrwi-^- 
Xi^itH^ E-'Z::>X''Hr{-f/>-t' 



The remaining MS. to be noticed is that which contains Hypoiid 
the oration of Hyperides against Philippides and the third p/„7/^,j,/. 
Epistle of Demosthenes (Brit. Mus. Papp. cxxxirr, cxxxivjl <^em. 
These two compositions, written on a single roll of papyrus, 
are in quite different hands, both of them rather peculiar 
and noticeable. The Hyperides is written in small, very 
graceful letters, well-rounded, with ligatures as distinctly 
marked and carefully written as the letters themselves, 
and in exceptionally narrow columns. The Demosthenes 
is in yet smaller letters — indeed there is no other literary 

and the Demosthenes as of the first century ; which would allow of the 
Bacchylides remaining where I have placed it above, at the transition 
from the Ptolemaic to the Roman style in the first century b. c. Messrs. 
Grenfell and Hunt further state that the forms of ^ and v which appear 
in the Bacchylidts are also found in two of their papyri of the Roman 
period, the Aristoxenus and Thucydides fragments ; but the /x in both 
these MSS. is quite different, being of only ordinary breadth and much 
more deeply indented, and the i;, though shallow in the top. does not very 
closely resemble the same letter in the Bacchylides. On the whole, the 
Oxyrhynchus papyri, which are all of the Roman period, seem to me to con- 
firm the date here assigned to the Bacchylides. Prof. Blass (Bacchylidis 
Carminu, 1898, pp. vii, viii) assigns the Bacchylides MS. to the first century 
after Christ rather than to the previous century, on grounds not of 
palaeography but of orthography, especially its comparative freedom from 
iotacism. His argument is that (according to the evidence of inscription.s) 
under the earliest emperors the interchange of ei and i was carried to an 
extreme, while in the course of the first century a reform in this respect 
was introduced, culminating in the second century under the influence of 
Herodian. In the absence, however, of more manuscript evidence on the 
point than is at present available, this argument is somewhat precarious. 
It is certain that in non-literary papyri the interchange of ei and 1 is quite 
as common in the first and second centuries after Christ as in the Ptolemaic 
period, and it is clear that much must have depended on the practice of 
individual scribes. A well-written MS., such as the Bacchylides, copied 
from an earlier MS. free from iotacism, would not be likely to l>e affected 
by it to any serious extent, certainly not to any extent comparable to 
inscriptions or non-literary papyri in which the scribe had no earlier copy 
to follow. Under these circumstances, the evidence of orthognipliy seems 
much too doubtful to be accepted as tlie main guide to the date of a MS. ; 
and the fact that tliis evidence leads Prof. Blass to assign the Herodas 
papyrus to the Ptolemaic age may be taken as reinforcing this conclusion 
(see below, pp. 94, 95 \ 

' Specimen facsimiles in Classical Toxts from Papyri in the British Museum, 
ed.F. G. Kenyon ,1891), plates ii and iii. Alphabets in App. I, cols. 7 and 8. 


papyrus in which the characters are so small — more angular 
and less well formed. The columns of both lean decidedly 
to the right. In detail the letters are akin to those of 
the Herculaneum papyri, and it is on the strength of 
this comparison that the MS. is assigned to the first 
century b. c. In the Hyperides, A has the loop in the 
left-hand corner which has been noticed in the Herculaneum 
papyri (and in an incipient form in the Louvre dialectical 
fragment), the right-hand stroke being made separately ; 
A is formed in a very similar way, but the loop does not 
fall below the rest of the letter ; € is well rounded, with 
the cross-stroke rather high ; A shows some projection 
of the right-hand stroke ; K is deeply curved, a sign of the 
approach of the Roman style ; Z is Ptolemaic, with three 
distinct strokes ; C is well rounded ; Y is deeply curved, 
much as in the reign of Augustus ; CO is fully and carefully 
Demo- formed. The Demosthenes has many of its details very 
Ep. III! similar, but is less well written throughout. A varies 
between the uncial form and the form with the loop in the 
left-hand corner; some shapes of the latter show the 
approach of the Roman type, but the letter is still written 
in two strokes ; € is small and narrow ; A sometimes has 
the right-hand stroke projecting, sometimes not ; M is 
generally shallow, recalling the Ptolemaic type ; Z has its 
three separate strokes ; C has rather a flat top ; Y resembles 
the same letter in the Bacchylides papyrus ; and CO is only 
slightly formed, the second loop being often hardly visible. 
It will be seen from a consideration of these features, and 
especially of such test letters as A, M, and Z, that both 
hands of this MS. are on the border line between the 
Ptolemaic and the Roman periods. Some forms are still 
distinctly Ptolemaic, while others are already approaching 
the types which will be seen in the next chapter to belong 
to the Roman style. They stand or fall with the papyri 
of Herculaneum, and cannot be placed very far from them 
in point of time. 


With these MSS., therefore, we reach the end of the 
Ptolemaic period, which may be roughly fixed at the 
middle of the first century b. c. Such precise definitions 
of time are, however, only useful as aids to the memory, 
and it is as well to repeat the caution that no precise 
accuracy can be expected in the dating of literary 
papyri in the present state of our knowledge. Palaeo- 
graphical periods necessarily melt into one another, and 
the forms of letters have no precise boundaries. Even 
where materials are much more plentiful, as in the minus- 
cule MSS. of the twelfth to the fifteenth centuries, it 
is only necessary to look at the published series of dated 
facsimiles from MSS. in the Bibliotheque Nationale, to have 
any rash dogmatism checked by the recognition of the fact 
that types of hands run on much further than is commonly 
allowed for. Dating MSS. by palaeographical indications 
alone is, to a considerable extent, a science of conventions. 
We can say that such and such a MS. is written in a hand 
which was prevalent in, say, the first century ; but we can 
much less confidently affirm that it was itself certainly 
written in that century. Still, it is only by provisional 
generalizations that science progresses ; and this is as true 
of palaeography as of physics or chemistry. The crystal- 
lization of isolated observations into formulas may lead 
to some mistakes, but it also leads to progress, by the 
confirmation or rejection of the formulas; and so this 
mapping out of the province of Ptolemaic palaeography 
will, it may fairly be hoped, be of some use for the 
arrangement and systematization of existing knowledge, 
and as an aid towards the assimilation of the knowledge 
which future discoveries may bring in. The criteria which 
have been indicated in the course of this chapter meet, 
it is believed, all the data which are now forthcoming, and 
arrange them in an intelligible series; but their final 
verification, or otherwise, must come from the evidence 
which the spade of the explorer has still to bring to light. 



The Roman period is, in relation to non-literary papyri, 
the period concerning which we have the fullest and the 
most precisely dated evidence ; but the same cannot be said 
of it in respect of literary MSS. In mere number, indeed, 
the literary documents of the Roman period exceed those 
of the Ptolemaic; but it happens that hardly any of 
them can be even approximately dated upon independent 
grounds. Consequently there has been great uncertainty 
as to the dates assigned to many of the MSS. with which 
we have to deal in this chapter; and it would be rash 
to say that the period of uncertainty is yet at an end. 
Any day some fresh piece of evidence may be brought 
to light which will modify the conclusions to which we can 
now come ; and all statements which rest solely on the 
impression produced by a MS. itself, without collateral 
evidence connecting it with better established knowledge, 
must be taken as only approximate. 

We have seen, at the close of the preceding chapter, 
the progress of the transition wliich leads from the 
Ptolemaic to the Roman age, and have found it exempli- 
fied cliiefly by the papyri from Herculaneum. Speaking 
generally, the characteristics which distinguish the Roman 
period from its predecessor are (i) a greater roundness and 
smoothness in the forms of letters, and (2) a somewhat 
larger average size. The very small hands which occur 


in so many — and those often the best written — of the 
Ptolemaic MSS. do not recur ; it is rather the characte^^s of 
such papyri as the Louvre Hyperides and the BacchyHdes 
that set the pattern for the Roman period. And letters 
which are prevailingly straight or angular in the earlier 
period, such as A, M, Z, are now rounded off into curves; 
while letters like C and C, which were rounded before, 
become more rounded now. 

Before coming to a definitely dated MS. which shows 
the characteristics of the Roman style fully established, 
there are one or two papyri to be mentioned, in which 
the remains of Ptolemaic influence seem to be visible. 
One of these is a fragmentary roll of the last two books Homer, 
of the Iliad, now in the British Museum V It is written xxiv!'"' 
in a hand of fair size, somewhat square in shape, with 
thin, firm strokes. The l adscript is regularly written, 
which is a sign of a relatively early date ; and there are 
no breathings, accents, or marks of elision by the first 
hand. There are, however, some of the critical symbols 
(the hi-nXi] and asterisk) employed by Aristarchus, of which 
this MS. provides the earliest examples as yet extant. 
In form several of the letters show distinctly Roman 
characteristics. A is formed in a single stroke, with 
a curved body, though there are also signs of the older 
angular shape : M is deepl}^ curved ; € and C are well 
rounded, the former having the cross-bar high ; the right- 
hand stroke of A projects considerably ; and Z is of the 
Roman type, formed continuously. On the other hand 
Y has the shallow top which appears also in the BacchyHdes 
and Demosthenes ; and this, along with the somewhat tran- 
sitional form of A and the generally early aspect of the 
writing as a whole, seems to justify the attribution of the 
MS. to the latter half of the first century b. c. 

' Pap. cxxviii; collation, with specimen facsimile, in Classical Texts from 
Greek Papijri ; complete text in Journal of Philology, xxi. 296. Alphabet in 
App. I, col. 10. 

i G 






Brit. Mus. 


Another MS., of less palaeographical, but greater literary, 
interest, which has sometimes been assigned to the Ptole- 
maic period, is the f ragment-of Alcman, now in the Louvre ^. 
As containing some otherwise unknown lines of this lyric 
poet, it is a papyrus of considerable value: but it is not 
a regular literary MS., being roughly written in a non- 
literary hand. This hand is, no doubt, a fairly early one, 
but there seems to be no reason for placing it earlier than 
the Christian era. All the test letters are of the rounded 
Roman tj^pe; while Y has the deep curve and the tail 
bending far to the right, which are characteristic of the 
non-literary hands of the first half (and especially the first 
quarter) of the first century. The fragment may be of 
the reign of Augustus, but, if so, it belongs to the last 
years of that emperor. 

We are fortunate in possessing a MS. which enables 
us to fix the exact state of palaeographical development 
which had been reached at the beginning of the Christian 
era. This is Pap. cccliv of the British Museum, a docu- 
ment of non-literary character, but written in a careful 
and most elaborate book-hand, and capable of being 
precisely dated ^. It is a petition for redress of injuries, 
addressed to the prefect of Egypt, Gains Turranius. This 
officer is known from an inscription (C. /. G. iii. 4923), 
which contains a date in the reign of Augustus ; and 
though the date is somewhat mutilated, and has been 
differently read by different scholars, it is certain that 
it corresponds to either 15, 10, or 7 B.C. The difference 
between these dates is quite immaterial for palaeographical 
purposes, and we thus have a definitely fixed date for the 
type of hand shown in this MS. It is a hand of great 
beauty, of medium size, regularly and firmly written, with 

' Facs. in atlas to Notices et Exiraiis, pi. 1. 

' Facs. in the atlas to vol. ii. of Greek Papyri in the British Museum, and 
a partial facs. (showing best-preserved part; in Plate XIV. Alphabet in 
App. I, coL II. 

Plate XIV. 

^l)^^f j;k)o^c:o e:>JiH^^ 

•A}^rnMA. CT )Q"rTOT/>-i>r ,'3-7 1 f xjMC^K-crj'cct:' >^ 'X^^/ i 

PETITION.— arc. B.C. lO. 



graceful and well-formed strokes and curves. There are, 
perhaps, no better written papyri in existence than this 
and the MS. to be mentioned next, which is closely akin 
to it. Of the individual letters, A is generally rounded, 
but with some examples of the angular shape ; M is pointed 
in the middle, which recalls one of the Ptolemaic forms ; 
Z is written continuously, but with more sharpness and 
precision than is usual in later Roman MSS. ; Y is 
moderately curved, and the tail is usually kept to the 
right, though not bent away as in the example last 
quoted; €, O, C, CO, are all well rounded and fully formed; 
the right-hand stroke of A projects upwards. 

It so happens that the discovery of this document The 
enables us to assign a date, with approximate accuracy, jiugguju 
to a literary papyrus of considerable interest, containing Odyssey. 
the latter part of the third book of the Odyssey'^. The 
interest of this MS. lies not only in the fact of its being 
the earliest extant copy of the Odyssey (which is of very 
much less frequent occurrence in papyri than the Iliad), but 
also in the beauty of the writing, which entitles it to be 
regarded as the handsomest literary papyrus at present 
known to exist. A teronhius ante quern is given by 
the occurrence of some scholia, written in a cursive hand 
which appears to belong to the end of the first century ; 
but a more precise date may be derived from its close 
resemblance in general appearance to the MS. which has 
just been described. The resemblance, which cannot be 
missed by any one who compares the two documents, 
renders it certain, even apart from the collateral evidence 
furnished by the scholia, that the Homer was written at 
no great distance of time from the petition to Turranius ; 
while an examination of details shows that the latter is 
the earlier of the two. In the Homer the A is more 
rounded, M is deeply curved, E is written continuously, 

' Brit. Mus. Pap. cclxxi. Facs. in Pal. Soc. ii. 182, paitial facs. in Plate 
XV ; text in Journal 0/ Philology, xxii. 238. Alphabet in App. I, col. 12. 

G 3 


and curves replace the sharp angles noticed in the petition. 
In all these typical letters the angular shapes have given! 
way to curves which are characteristically Roman in 
nature. Of the other letters, € and C are well rounded, 
the former having the cross-stroke very high, much as 
in the Iliad MS. last described ; H has the cross-bar ex- 
ceptionally high ; and Y approaches the y-shape, the two ] 
upper limbs branching off more equally than has usually i 
been the case hitherto. This is a transition shape in 
the direction of the V-form which prevails later. The! 
general conclusion, therefore, to which palaeographical 
data point is that the Odyssey papyrus was written near 
the beginning of the first century, and that the scholia J 
which are of later appearance and include quotations 
from the grammarian Apion, who flourished under 
Caligula, were added in the second half of the same 
The More doubt; attaches to another Homer papyrus, con- 

Hams taining the eighteenth book of the Iliad, and commonly 
known (from the name of its first purchaser) as the 
Harris Homer \ This was one of the first literary papyri 
to be brought to light, and consequently enjoyed a some- 
what exaggerated reputation of extreme old age. Sir 
E. Thompson, whose authority in such matters is of the 
highest, places it ' without hesitation as early as the first 
century b. c.^.' The increase of evidence during recent 
years makes it wery doubtful whether so early a date 
can be maintained. The whole aspect of the hand is 
Roman, not Ptolemaic, and one might even be inclined 
at first sight to refer it to the second century, were it 
not for certain letters which still recall the hands of 
the first century b. c. These are especially A, which is of 

' Brit. Mus. Pap. cvii ; faes. in Pal. Soc. ii. 64, and a less successful one 
in the Cat. o/Anc. MSS. .Greek). Alphabet in App. I, col. 13. 

^ Handbook, p. 124. In the Pal. Soc. and Cat. of Anc. MSS. it is simply 
described as 'perhaps first century b. c.*; but the Handbook is the latest 

Plate XV. 

Vpcth^'-rkf ktX)0^xxu yxx<l).olc>^oi ciM^ 

OM K)<kr on oroy Cog exo y-c cnx oi^ 
. 4 i^eMrM<|)XpocK-xxdN:Ex\eMnAexi'-rt 


P XxiNYN 60X CMC Noitui >^N c? cccceyc 
: > o:iM M^ijo^M 0X0 eTNfTi! oe?4i Kf7 ceo i C-> 
. ■XyrxtCTrCinoci ocKAieA^TYOoexeforo( 
^l^ TO [q AC-U-Y agj N^n ?x ecetnj^u ocrirn ?j:ii 
I •. n Ja kec^XiOi Arerrt x tia ioJu/t r xxxptt^ XX ci 

: a;kj ii^iknx cAu xxocn^f^rxxxextn cerrj a 

^ fftf^CiijcoNTaiiol^xrion^^^ 

i| ^^c^4Y)cTCecx^J^x-l^£S^^ 
"l^iici^&^P'i rCN^ ct x^.XMnf a:c ix^n;^> '- rt- 





the angular type which is found in the Bacchylides, though 
somewhat less sharply pronounced; and 6, which, again 
as in the Bacchylides, is narrow, with a long projecting 
cross-bar, and with the head of the letter often separated 
from the rest. On the other hand M and Z are distinctly 
Roman in character ; A always has a projecting right-hand 
stroke; M, N, n, are square, not elongated as in Ptolemaic 
hands ; C has a somewhat flattened top ; Y is approaching 
the V pattern, and has quite lost its Ptolemaic appearance. 
Some omitted lines have been supplied in the margin in 
an ordinary cursive hand which seems to belong to the 
beginning of the second century ; and it is probably the 
same hand that has added the accents and breathings. 
On the whole it would appear that the MS. must be 
assigned to the first century of our era ; though of course, 
here as always, no pretence of dogmatism is admissible, 
especially in conflict with other authorities. 

Another MS. which has commonly been placed in the The 
first century b. c, and sometimes even earlier, but which iji'Jg^'^„',j 
may now almost certainly be brought down to the first Hypor- 
century after Christ, is the great papyrus of Hyperides — 
the one by which his work was first made known to the 
modern world — containing the prosecution of Demosthenes 
and the defences of Lycophron and Euxenippus — the 
latter being the only extant speech of the orator which 
is absolutely intact. At the date when this MS. was 
acquired by Messrs. Harris and Arden, there was practically 
no means of forming an accurate idea of its age, and its 
first editor, Mr. Babington, cautiously placed it between 
the second century before, and the second century after, 
the beginning of the Christian era. Subsequent scholars, 
who have tried to be more precise, have in fact ranged 
over the whole of this somewhat wide period. Sir 
E. Thompson has assigned it to the first century B. c.^ ; 

1 Cat. 0/ Anc. MSS. {Greek) ; Handbook, p. 123. A complete facsimile 
of the MS. (by hand) is given in Babington's edition ; photographic 


Prof. Blass to the second century after Christ ^. But 
this discrepancy between two eminent palaeographers 
is due to the fact that they based their conclusions upon 
different evidence. Prof. Blass was guided by the cursive 
hand in which the title of the MS., giving a list of the 
orations contained in it, is written, which he rightly 
assigned to the second century after Christ ; while Sir 
E. Thompson equally rightly argued that the title is 
not in the same hand as the body of the MS., and there- 
fore gives nothing more than a termitius ante quein 
for our assistance. It is, however, now sufficiently clear 
that no date within the Ptolemaic age is possible for 
this style of writing. The most marked feature of the 
hand in its general appearance is its roundness — a dis- 
tinctively Roman characteristic ; and the forms of the 
single letters all point the same way. A is written con- 
tinuously, with rounded angles and somewhat loose con- 
struction ; M is fully rounded ; Z is continuous, and rather 
irregular ; C is rounded, and even shows a decided tendency 
to fall forward — an indication of a date nearer the end of the 
first century than the beginning ; Y is of the characteristi- 
cally Roman type. The letters are not always well or firmly 
formed ; and that again is a sign of a relatively late date. 
On these grounds alone it would seem almost certain 
that this papyrus cannot be placed earlier than the second 
half of the first century ; and this conclusion has been 
strongly confirmed by evidence which has lately come 
to light. In the first place, striking resemblances to 
this Hyperides MS, are found in the long poll-tax rolls 
recently acquired by the British Museum, which bear 
dates in the years 72 and 94^. Secondly, a small scrap 

facsimiles of parts of it in Pal. Soc. i. 126, and Cat. nf Anc. MSS. Some 
recently discovered fragments, now in the possession of Rossall School, 
are shown in Plate XVI, through the kind permission of the Kev. J. P. 
Way, head-master of Rossall. Alphabet in App. I, col. 14. 

' I. von Miiller's Handbuch der klassischen Altertwms-Wissenschaft, ed. 2, p. 312. 

* Brit. Mus. Papp. cclvii-cclix, cclx-cclxi. 

Plate XVI. 



i^!iv; 4,:,- vix-x ;::,!••■•' - 

■;*\;! itP 






•7,'!". "^ 




J - -^ 




of a literary papyrus, in a hand almost identical with 
the Hyperides, but somewhat larger, was acquired along 
with a number of dated documents of the first and second 
centuries^. But the most conclusive evidence is afforded 
by a fragment brought from Egypt by Mr. Grenfell 
in 1H95, and now in the Bodleian Library. It contains The Bod^ 
a small portion of the Homeric Lexicon of Apollonius, Apoiio- 
thouo'h in a form somewhat dift'erent from that in which "^"^- 
it has hitherto been known ^. The evidence for its date 
is a little complicated, but seems fairly conclusive, at 
least as regards the terminus a quo. The grammarian 
Apion is mentioned in various places in the Lexicon, and 
one of these passages occurs in the Bodleian fragment. 
The papyrus is, however, mutilated in this place, and 
the name of Apion is not preserved; so that it is con- 
ceivable — considering the extent to which the texts of 
Apollonius hitherto extant have been expanded from the 
original — that the name might be a later addition. But 
the size of the lacuna just suits the supposition that the 
papyrus here had the same text as the later copies, and 
there is no other natural way of filling it. It is therefore 
a legitimate conclusion that Apion was mentioned in the 
papyrus, when it was perfect ; and since Apion flourished 
under Tiberius and Caligula, it is practically certain that 
the papyrus cannot be earlier than the middle of the 
first century, while it may, of course, so far as this argu- 
ment goes, be considerably later. The importance of this 
conclusion lies in the fact that the fragment is written 
in a hand of precisely the same type as the long 
Hyperides MS. ; and if the Apollonius is not earlier 
than the middle of the first century, the same must also 
be the case with the Hyperides. The likeness between 
the two MSS. is as great as it can be between two 

' Brit. Mus. Pap. ccviii c. 

''■ A facsimile has been privately circulated by Bodley's Librarian, to 
^vhom the identification of the fragment is due. 


liands which are not identical, and will Le disputed by 
no one who has compared the two. 

It seems established, therefore, that the theories which 
assigned the Hyperides to the Ptolemaic age, or even to 
the first century of the Roman period, must be abandoned ; || 
but there still remains the possibility that Blass is right 
in placing it as late as the second century. But, in the 
first place, the main ground of his view — namely, that 
the title of the MS. is written in a cursive hand of the 
second century — is unsound ; for the date of the title does 
not necessarily establish the date of the MS., but merely 
gives a terminus ante queni. Titles not only might be, but 
actually were, written much later than the MSS. to which 
they applied. A striking instance is given by the Bacchy- 
lides papyrus, in which some of the poems are without 
titles, while those which exist have plainly been written 
at a later date than the poems themselves. So far, then, 
as the evidence of the title goes, all that can be said is 
that the MS. cannot be later than the end of the second 
century, though it may be considerably earlier. It is here 
that the evidence of the poll-tax rolls, above mentioned, 
becomes important ; for they show that hands of the same 
class as that of the Hyperides were in use in the last 
thirty years of the first century, while no example of such 
hands has yet been found of a later date. To this period, 
then, of the end of the first century, the great Hyperides 
MS. must almost certainly be referred. External testimony 
here strongly supports the indications of palaeographical 
science, and it will need clear evidence to overthrow their 
joint conclusion. 
Brit. Mus. ^t this point in the history of papyrus-palaeography 
we come upon another document which, though not 
literary in its contents, is yet, like the petition to Turranius 
of circ. B.C. lo, of considerable value for the study of 
literary hands. It is a lease of an oliveyard and other 
lands in the village of Socnopaei Nesus, dated in the year 

Plate XVII. 






i 1. 

; ./ 

> r 

» ' V,I 


88 ; and instead of being written, as usual, in a sinall 
and very cursive hand, it is executed in careful uncials 
of a large size, with only a sprinkling of minuscule forms '. 
In size, the characters of this MS. are larger than any 
that we have yet described, though some later documents 
perhaps equal them. In type, they are plainly the fore- 
runners of the handsome uncials in which the earlier vellum 
MSS. are written. Before the discovery of this papyrus, 
with its precise date near the end of the first century, 
there was little or nothing to show that such hands went 
back earlier than the fourth century at most. The present 
document, though by no means perfect in execution, shows 
that we must look much higher for the origin of this 
hand ; and it also suggests a conclusion to which we shall 
have to refer again presently, that the palaeography of 
Greek papyri anticipated in its development the subsequent 
history of writing upon vellum, so that the corresponding 
stages of writing on the two materials are not contemporary, 
but are separated by some centuries of time. 

The letters of this papyrus are not uniformly uncial in 
character, some of them being merely the current minuscule 
forms written on a large scale. This is notably the case 
with A and Y, while €, which is correctly formed as a general 
rule, occasionally lapses, through the forgetfulness of the 
scribe, into a completely minuscule form. Of the other 
letters, M is conspicuously shallow in its central depression ; 
Z is of the regular Roman type, and somewhat cursive 
in appearance; C is well formed and upright; O is large 
and prominent ; CO is well written. The MS. as a whole 
must be regarded, not as a thorough-going example of 
a book-hand of a large uncial type, but as an indication 
of what the contemporary book-hand might be. We cannot 
say that this is a hand of the same character as those 

' Facsimile in Pal. Soc. ii. 146, and in the atlas to Greek Pavyrl in the 
British Museum, vol. ii ; partial facs. in Plate XVII. The papyrus is known 
as Brit. Mus. Pap. cxli. 




which are found in the great vellum MSS. of the fourth 
and fifth centuries ; but we can say that, if this hand 
could be found in a non-literary document of the year (S8, 
then a hand of the same character as those of the vellum 
MSS. may have been in use for literary papyri at the 
end of the iirst century. The large uncial hand, in short, 
instead of being a creature of the fourth century, called 
forth by the adoption of vellum into general use, exists on 
papyrus at least three centuries earlier, 
lirit. Mas. Further evidence in the same direction is given by another 
papyrus in the British Museum (Pap. CLXXXVii), which 
contains a small fragment of a moral or historical treatise, 
giving a description of the training of the youths of some 
nation, apparently the Lacedaemonians ^. Only some sixteen 
lines are preserved intact, or approximately so, with small 
traces of some other lines in the same and adjoining 
columns ; but quite enough remains to show that it was 
a carefully written MS., in a large uncial hand. The letters 
are of about the same size as those of the lease just 
described, and are free from the intermixture of minuscules 
by which that is disfigured. Its date is not known with 
certainty, but the evidence which has just been given 
authorizes us to place it at the end of the first or in the 
course of the second century ; and this is confirmed by 
the fact that the verso of the papyrus bears writing in 
a cursive hand which may safely be referred to the 
beginning of the third century. The writing of the 
AaKebaifj-oviMv TroAtreia (or whatever the treatise is) is careful, 
delicate, and handsome ; and its size and shape put it 
definitely in the line of ancestry of the vellum uncials. 
A continues to be of the rounded form ; M is very deeplj^ 
curved ; C is so much rounded as to be hardly distinguish- 
able from O, a feature which is perhaps in favour of 
a date in the second rather than the first century. The 

* Published (without facsimile) in the Rcv^ic de Philologie, xxi. i (1897). 


columns of writing are narrow, containing, in this large 
hand, only some twelve to fifteen letters. 

Between the first and the second centuries after Chi'ist 
it is impossible, at any rate with the evidence now available, 
to draw any firm line of palaeographical demarcation. No 
external change came over the government of Egypt which 
could affect the local handwriting. The development of 
the Roman hand pursued a natural course ; and in the 
absence of precisely dated examples it is impossible to 
fix its stages with certainty. There are a considerable 
number of MSS. which appear to belong to this period ; 
but it would require a bold palaeographer to decide, with 
regard to most of them, whether they were written in 
the second century or the last half of the first. It so 
happens that most of the MSS. referred to contain parts 
of Homer; but the list also includes the unique papj^rus 
of Herodas, a speech of Isocrates, and a few minor 
pieces ^. 

Close to the border line, but almost certainly within The 
the limits of the first century, comes a papyrus of very j^j^^yurn 
great literary interest, namely that which contains the Ai-istotle. 
'A6r]vaLoiv rJoAtreta of Aristotle ^ ; but unfortunately, of the 
four hands in which it is written, only one at all approaches 
the character of a book-hand, and that is plainly not 
the work of a trained scribe, from the number of clerical 
blunders which it contains. Rather is the MS. a striking- 
example of what may be called the private method of 
circulating literature in the ancient world. Libraries were 
few and far between, and copies of literary works written 
by professional scribes may have been difficult and expen- 
sive to obtain in the upper parts of Egypt. Hence we 
seem to see the existence of a practice of lending MSS. 

' The list has been considerably increased by the recent discoveries 
of Messrs. Grenfell and Hunt at Oxyrhynchus, which include several 
literary fragments belonging to this period. 

^ Complete facsimile published by the Trustees of the British Museum 
(1891) ; specimen plate in Pal. Soc. ii. 122. 


to be transcribed privately by those who wished to possess 
copies of the work ; and if new papyrus could not be 
readily obtained, a roll which had already been used 
might be made to do duty again. This is what happened 
in the case of the Aristotelian treatise. It is written on 
the back of some old farm accounts ; and the greater 
part of it is written in hands of a very cursive character. 
Apparently speed was necessary — perhaps the MS. from 
which it was copied could only be lent for a short time ; 
for the different scribes evidently began independently, 
and their several sections do not exactly join. Two of 
the writers seem to be men of culture, able to write Greek 
correctly ; and since their hands are similar in appearance 
(though certainly not identical, as has been suggested), 
it is perhaps not very fanciful to suppose that they are 
those of the owner of the MS. and a near kinsman. The 
other scribes seem to have been slaves, and their Greek 
is far from correct. One of them writes a sort of rough 
uncial hand, square, thick, and inelegant ; the other, who 
takes up his work when he has finished it, tries at first 
to imitate his uncials, but quickJy slides into a loose 
and ill-formed cui'sive hand, in which he completes his 
portion. Some of the more glaring blunders of the illiterate 
scribes are corrected in the hands of their educated 
collaborators.. The date of the whole MS., is limited on 
the one side by the fact that the farm accounts on the 
recto are dated in the years 78-9; while the close resem- 
blance of the cursive w^riting of the Aristotle to that of 
several documents (since discovered) of the reign of Domi- 
tian ^ makes it higlily probable that it was not written 
very far from the year 90. 
The auto- A somewhat special interest attaches to this MS., not 
fhe^New o^^^y on the ground of its contents, but also because 

Testa- the method of private transcription, of which it is an 
rnent. ^ ^ 

' See above, Plate VI, for an example. Another (of a. d. 83-4) is given in 

the Fiihrer durch die Ausstellung der Papyrus Erzherzog Bainer, pi. ix. 


example, must have been very commonly employed in 
the case of the Scriptures of the New Testament. The 
Christians of the first century were not predominantly 
an educated body, nor were they, as a rule, wealthy. 
Hence, in many parts of the Greek world, they could 
neither transcribe their Scriptures in the best literary 
hands of the period, nor pay professional scribes to do 
so for them ; and they must consequently have often 
fallen back on the method of amateur, or private, tran- 
scription. This would especially be the case in times of 
persecution, when the trade of transcribing the Scriptures 
could not have been carried on publicly even in the 
communities where there were otherwise facilities for 
such work. In the writing of the Aristotle, therefore, 
one may see an example of the manner in which the 
Christian Scriptures were often transmitted ; and in these 
circumstances one may find part of the explanation of the 
early rise of a large number of various readings. No 
doubt there were exceptions, and some of the autographs 
of the New Testament books may have been well and 
carefully written ; but the manner of transmission just 
described must have been employed on all the books 
in some stage of their history, and its characteristics 
should therefore be borne in mind as a datum in the 
textual criticism of the New Testament. 

This, however, is somewhat of a digression from the 
immediate subject in hand, the development of the literary 
form of writing on papyrus ; and it is time to return 
to the other MSS. which, as mentioned above, seem to 
fall within the latter part of the first century or in the 
course of the second. One of these, which ranks next 
to the Aristotle in point of literary interest, is also 
remarkable as standing rather apart from the general 
line of development, and provides a salutary proof of the 
error of assuming that only one type of hand can have 
prevailed at any one period. This is the papyrus of 


The Herodas'^, containing the mimes of that otherwise prac- 

Papjius. tically unknown writer of the Alexandrian age. As has 
been stated in an earlier chapter, it is the smallest papyrus- 
roll extant in respect of height, measuring no more than 
5 inches, and containing only i8 lines, on an average, 
in each column. As an almost necessary consequence, it 
is written in a small and compact hand, clear and easy 
to read, but almost entirely devoid of ornamentation. 
Indeed the matter-of-fact realism of the poems seems 
to be reflected in the plain, unadorned appearance of 
the writing. The forms of the letters, when examined 
in detail, are unquestionably of the Roman period, but 
the general appearance of the writing is so unlike that 
of any other extant papyrus, that it is exceptionally 
difficult to fix its date from palaeographical considerations. 
In the editio princeps it was assigned to the second or 
third century, but increased knowledge makes it almost 
certain that this date is too late. The clearest proof 
of this comes from the accuracy with which non-literary 
hands can now generally be dated. In the course of the 
MS. a line has been accidentally omitted, and has been 
supplied in the upper margin of the papyrus in a cursive 
hand ; and it happens that this contains an 77 ctf the pecu- 
liar form (^) which has been described in a previous chapter 
as characteristic of the period from about A. d. 50 to 1 60. 
It is, no doubt, impossible to say how long after the 
transcription of the MS. this omission was made good. 
It may be said that the probabilities are against an error 
in so comparatively rare a poet, and in a MS. in private 
hands, being corrected from any except the original MS. 
from which it was copied ; but such a probability does 
not amount to an argument of much strength. It is, 
however, clear that the third century, and even the last 

' Bi"it. Mus. Pap. cxxxv. Complete facsimile published by the Trustees 
of the British Museum (1892) ; text, with specimen facsimile, in Classical 
Texts ; facs. of one column in Plate XVIII. 


Plate XVIII. 

- 1-1 n ^ re9 or2^ Z-XLt ^>o A> ,'rt «-^ rJ^- cf^ W 4> V 



part of the second, is too late a date to assign, and that 
the MS. should rather be placed in the first century 
or the first half of the second century. The forms of 
the individual letters require little notice; A is of the 
rounded type ; M is deeply curved (ju, u) ; Z has the top 
stroke separate, but the middle and lower sti'okes united 
(IE ), a variant which imiay indicate a relatively early date, 
but is so rare as to provide no secure basis of argument ^ ; 
Y is very stiff and straight, usually with a very short 
tail. Throughout it is a plain representation of Roman 
semi-uncial, with less grace than usual, but quite without 
affectation or mannerism ^. 

Another MS. which should probably be placed about Tlie 
the turn of the century is the British Museum papyrus Museum 
of Isocrates Ilepl Etpi/z;?;?^. It is a large papyrus, mutilated lsf>crates. 
at the beginning, and written in two distinct hands, the 
first of which is decidedly better than the second. The 
latter indeed verges closely on the non-literary type of 
hand, and is certainly not the work of a trained scribe. 
The first hand has several features in common with the 
large Hyperides MS. A is rounded; Y is curved, not 
forked, which favours a date in the first century ; C falls 
forward to some extent, which shows that it cannot be 
early in that century. The other letters are not distinctive. 
The second hand is loose and straggling to an extent 
which is not found before the latter part of the first 

^ This form recurs in the Oxyrhynchus fragment of Demosthenes, men- 
tioned above (p. 76). 

^ Prof. Blass has recently, in a sort o? obiter dictum, assigned the Herodas to 
the Ptolemaic period {Bacclujlidis Carmina, pp. vii, viii), on the ground of its 
frequent interchange of i and ei, which he regards as characteristic of this 
period ; but (i) a study of the non-literary papyri of the first and second 
centuries shows that such iotacisms were extremely common then (and 
this evidence is especially applicable to a MS. which, like the Herodas, is 
evidently not the work of a highly-trained sci'ibe) ; and (2) the forms 
of the letters are wholly of the Roman type. 

^ Brit. Mus. Pap. cxxxii. Two specimen facsimiles in Classical Texts, 
with collation. 



piipyri : 
Brit. Mu.- 

century at earliest. Its A, C, and Y are substantially the 
same as those just described ; and it has a very cursive 
form of e. On the whole, the end of the first century 
or the first half of the second seem to be the most 
probable dates for this MS., which, it may be remarked, 
is of some importance in the history of the transmission 
of Greek literary texts, since it shows that, of the two 
well-marked families into which the vellum MSS. of 
Isocrates are divided, neither is exclusively to be trusted, 
and that the formation of these families does not go back 
to classical, or even to early Christian, times. It also 
tends to show that the text of Isocrates was suhstantially 
the same in the first century of our era as that which 
has been preserved for us in MSS. a thousand years 
later. But this is to trespass outside the domain of 

We come now to a large group of Homer MSS., including 
three somewhat small fragments in the Louvre, three long 
rolls in the British Museum, and a long and very inter- 
esting MS. in the Bodleian. It would be absurdly rash 
to try to place these in a precise chronological sequence ; 
but it is worth while to describe them severally, and to 
show on what grounds they are assigned to the period 
now under consideration. 

In one instance there is external evidence in the shape 
of dated writing on the other side of the papyrus. This 
is Brit. Mus. Pap. cxxxvT, a roll containing nearly the 
whole of the fourth book of the Iliad, and fragments 
of Book III \ It is written in a rough, uneducated hand, 
which gives it an appearance of later date than seems 
to be actually true. The recto of the papyrus is occupied 
by accounts, and among these are dates which must belong 
to the reign of Augustus ; and since a papyrus of accounts 

* Specimen facsimile ia Classical Texts, Plate vii. The MS. is there 
wrongly dated, the evidence derivable from the writing on the recto not 
having been discovered. Attention was first called to it by Prof. Wilcken. 

Plate XIX. 


'"cf^^rffi -^^ ^^ rX-^ 0>fBJv«;r'-:!t^-^^"^^^'^'^ V?-; '^^-'^ 




would not be likely to be preserved very long (the case 
of- the Aristotle MS. shows how soon such a papyrus 
might be used again), it seems almost certain that the 
Homer must have been inscribed upon the verso in the 
course oi' the first century. On the other hand the over- 
hanging shape of C and the v-shape of Y make it impossible 
to place it before the latter part of the century; and it 
may be even later than this. It is, however, so plainly 
not the work of a professional transcriber of literary 
MSS., that its importance in palaeographical history is not 

Much more interesting, from the palaeographical point Brit. Mus. 
of view, is a papyrus brought from Egypt by Mr. Grenfell ^^^^^ 
in 1896, and now in the British Museum (Pap. Dccxxxri)^. 
It is one of the longest Homer papyri in existence, 
containing the greater part of Iliad xiii and xiv, carefully 
written in a fine and ornamental hand. Indeed in point 
of appearance it is perhaps inferior to none but the British 
Museum Odyssey papyrus, described above. It has the 
uncial form of A — always a sign of a carefully written 
MS., but not confined to any one period ; and the general 
aspect of the hand is square, firm, and well-formed. The 
other letters have the shapes which we have seen to be 
characteristic of the fully-developed Eoman hand; € is 
well-rounded, with rather a long cross-bar; Z is square 
in form, and its cross-stroke not unfrequently meets the 
lower bar in the middle instead of at the end, which 
is an archaic feature ; H has its cross-bar very high ; 
the right-hand stroke of A projects but little above the 
point of junction; M and Z have their regular Koman 
forms; C is well rounded, but upright; Y is v-shaped; 
S' is very angular ; and CO is rather stiff" and compact. 
The whole writing is rather compressed, and has a ten- 
dency to lean backwards. So far as the forms just 

 A facsimile of a portion is given in Plate XIX. Alphabet in App. I, 
col. 15. 









described differ from the average Roman hand of the 
end of the first and the second century, they lean towards 
the earlier forms ; and consequently, if one Avere compelled 
to assign a date within narrow limits, one would be disposed 
to place it in the first century. It is, however, just a case 
where additional care on the part of the scribe may make 
a MS. appear older than it is, through his retaining the 
older forms of letters ; and it would consequently not be 
wise to attempt to define the date too closely. 

The same may be said of another Homer-papyrus, which, 
in contrast to that just described, was the earliest of its 
class to be discovered. This is the MS. known as the 
Bankes Homer, from the name of the gentleman who 
brought it from Egypt in the year 1821. It contains 
the last book of the Iliad, in very good condition^. It 
is written in a typically Roman hand, but with less 
precision and compactness than the last example. A is 
of the angular form which is seen as far back as the 
first century e. c, but is more irregularly written ; the cross- 
bar of € is low; A has a long projection above the point 
of junction; C is round and upright; Y is v-shaped and 
rather straggling. The greater looseness and irregularity 
of the writing suggest that it is somewhat later than 
the last-mentioned MS., and that it should be placed in 
the second century, as it has been by Sir E. Thompson; 
but here again one does not know how much to allow 
for the personal equation of the scribe. There are a few 
corrections in a cursive hand which appears to be of 
this century, but one cannot tell how long after the 
original MS. they were written. An earlier date than 
the first century, such as used formerly to be ascribed 
to this MS., may be pronounced quite impossible in view 
of modern knowledge. 

The Homer papyri of the Louvre may be dismissed 

' Now Brit. Mus. Pap. cxiv ; specimen facsimile in Cut. of Ancient MSS. 
{Greek}, plate vi, and in Pal. Soc. i. 153. Alphabet in App. I, col. 16. 


more svimmarily, being very much smaller in extent The 
and of no importance, either palaeographical or textual. Uomors 
Three fragments are included in the publication in the 
Notices et Extraits (nos. 3, 3''^®, and 3^")^ Of these the 
best from the palaeographical point of view is no. 3**''", 
a small fragment of what must once have been a hand- 
some copy of the thirteenth book of the Iliad. It has 
the uncial A, but all the other test letters, such as M, Z 
and Y, are distinctly Roman. It may belong to either 
the first or the second century. Papyrus no. 3'''^ is less 
handsomely written, and has the rounded minuscule form 
of A ; but there is nothing in the shapes of the letters to 
show that it is substantially later than the others. The 
remaining fragment, no. 3, is the least ornamental of the 
three, and probably the latest in date, being written in 
a somewhat loose and shaky hand. It does not appear, 
however, to be later than the second century, and shakiness 
of hand is less to be relied on as a sign of a relatively late 
date in papyri than in vellum MSS. ; for the latter are 
almost invariably the work of professional scribes, and are 
generic rather than individual in their styles, whereas 
a large number of papyrus MSS. were written by amateur 
or semi-amateur transcribers, and reflect the idiosyn- 
crasies of the individual as well as the tendencies or 
fashions of the age in which he lived. In cases such as this, 
therefore, we must be content with a rather wide range of 

The first and second centuries represent the prime of the 
Roman style, and the principal MSS. belonging to them 
have now been described. It will have been seen that it is 
normally a hand of moderate size, with letters square in 
shape and well-rounded ^ ; moreover, so far as we have yet 

* Facsimiles (by hand) in tlie atlas accompanying vol. xviii. of N. et E., 
plates xii and xlix. 

" The terms are not contradictory, ' square ' implying that the letters, 
whether angular or curved in outline, may be enclosed within a square. 

H 2 


gone, the letters are upright, even though the columns 
of writing sometimes lean to the right. It is a neat, 
orderly, business-like hand, and in its best manifestations 
graceful and, so far as is consistent with simplicity, 
ornamental. It is very probable that a fragment of the 
Orestes of Euripides, now at Geneva, should be added to 
the list of beautifully- written MSS. of this period ; but 
its editor^, Prof. Nicole, who describes it as the handsomest 
example of writing upon papyrus known to him, does not 
fix its date precisely, and no facsimile of it has been 
published. Certainly, in the present writer's judgement, 
the best examples of the Roman period excel those of any 
other period in grace and beauty, as well as in general 
workmanlike qualities ; and this without in the least depre- 
ciating the attractiveness of the strength and precision 
seen in the best Ptolemaic papyri, such as the Louvre 
H3^perides or the British Museum Bacchylides -. 

Ptolemaic letters are generally broader than they are high, and Byzantine 
higher than they are broad, while Roman are normally about equal in 
each dimension. And whereas Ptolemaic letters are angular, Eoman have 
their curves well rounded. 

' In the Bevue de Philologie, xix. 105. 

- The account which has been given of the palaeography of the first and 
second centuries is confirmed by the recently-published Oxyrhynchus 
papyri. The published facsimiles from these papyi-i include a small but 
beautifully-written fragment assigned to Alcman, apparently of the end 
of the first century ; a fragment of Thucydides (Book iv.) in a small, 
irregular hand, probably of the latter part of the first century, but 
recalling in its size the much earlier papyrus of the third Demosthenic 
epistle (p. 78 above) ; and the fragment of the Upoolnia ArjuTjyoptKa of 
Demosthenes, to which reference has been made above in connexion with 
the Bacchylides MS., written in a broad, slightly sloping liand, rather 
loose in formation. Some other fragments of later date will be mentioned 
below. Another recently-pxiblished papyrus of this period is a mathe- 
matical fragment known as the Ayer papyrus, brought from Egypt in 
1895, and now in the Field Museum at Chicago (edited, with facsimile, 
by Mr. E. J. Goodspeed in the American Journal of Philology, xix. 25^ It is 
written in an ornamental hand approaching the uncial type, and has 
been assigned by some authorities to the second or even the third century, 
but would appear from the facsimile to be not later than the first. Some 
of the letters, especially Y, are quite of the early Roman type. 


Plate XX. 










. P ft; ^.y -^'*^ '^■^ ^-^?i ^'^1 





,-- c 

-* -■ ■-- '» :t„. ~1 -13 __ i. >'  »! 

^::%:> f i ^' ^ 

p- V* ..'^ t. 1 - '^ t1j v^ X ^,? 






' ^ ^'^ 



One MS. has, however, been reserved for separate con- 'i'^ie 

sideration, whicli has a strong claim to be regarded as yet 
handsomer than any that has hitherto been mentioned. 
This is a copy of the second book of the Iliad, found 
at Hawara by Mr. Flinders Petrie, and now in the 
Bodleian Library ; and it has been reserved on account of 
the doubt attaching to its date. Sir E. Thompson, whose 
judgement in palaeographical matters stands unquestionably 
first, originally assigned it to the fifth century, on account 
of its resemblance to the vellum MSS. of that date ; and it 
is with great diffidence that a different opinion is here 
propounded. The MS. is written in an exceptionally large 
and carefully-formed hand^. The letters are square in 
build, upright in position, with well-rounded curves. A is 
of the uncial shape, with the cross-bar rather high ; € has 
the cross-bar rather low, as also has H ; M is deeply curved ; 
Y is Y-shaped, but not of the Ptolemaic or early Roman type. 
The other letters possess no distinctive characteristics, but 
are square, well-curved specimens of the best Roman hand. 
Now in many respects the characteristics here described 
apply to the vellum uncials of the fifth and sixth centuries, as 
they are seen in such MSS. as the Codex Alexandrinus and 
the Codex Claromontanus ; and it is on this ground that 
Sir E. Thompson assigns to the papyrus a date contempora- 
neous with these MSS. On the other hand it will be seen 
that, except in point of size, the forms of the letters are just 
those of the Roman hand of the first and second centuries. 
At the time when the papyrus was discovered, no uncials 
of this size had been found among the papj^ri, and it was 
perfectly natural to associate them with the vellum uncials 
to which they unquestionably bear a considerable resem- 
blance. The position has, however, since been altered by 
the discovery of such documents as the lease of a. d. 88, 

^ Facsimile of one column in Petrie's Hmvara, BiaJimu, and Arsinoe', 
part of a column in plate XX. Alphabet in App. I, col. 17. 



described above, and the fragment of the AaKehaiixov'nav 
Ylokuda, which show that a large uncial hand was in 
use upon papyrus as early as the first or second century. 
Hitherto the genesis of the hands seen in the early vellum 
uncials has been unknown ; but it now begins to seem 
probable that they are the direct descendants of the hand 
which had been used for the best papyrus MSS. two 
hundred years earlier. Further reasons in support of this 
view will be found below, and will be considered in the 
next chapter, in connexion with the transition from 
papyrus to vellum ; but meanwhile there is another argu- 
ment in favour of an early date for the Oxford Homer. 
A few notes have been added in the margins by a hand 
different from that of the text; and these are written 
in a sloping hand which will be seen presently to belong to 
the third century \ No doubt a sloping hand also followed 
the vellum uncials of the fifth and sixth centuries ; but 
it is to be remembered that after the beginning of the 
seventh century Greek writing upon papyrus becomes very 
much less common, and there is nothing Byzantine in 
the hand in which these notes are written. 

It is not maintained that these arguments are conclusive, 
especially as the history of large uncial writing on papyrus 
is not by any means fully known to us; but they seem 
to establish a strong probability that the Homer may 
be placed as early as the second century. Some slight 
confirmation of this conclusion may be found in a com- 
parison of this MS. with the two minute scraps which 
appear at the bottom of the plate containing the fragments 
of the large Hyperides MS. (plate XVI). These scraps, 
coming from what must have been a very handsome copy 
of Demosthenes' Olyntkiacs, were found in a dunmiy 
roll of papyrus along with the Hyperides fragments and 

' One of these notes may be seen in the right-hand margin of the 
facsimile ; but it is so faint in the plate as to be of little use to any one 
who has not seen the original. 


other material, none of which was later than the second 
century. There is therefore some slight presumption 
that they were themselves not very much later than that 
period ; and it will be seen that there is a considerable 
resemblance between them and the Oxford Homer. It 
may be doubted also whether so handsome a copy 
would have been written on papyrus at a time when 
the substitution of vellum for that material had already 
been in progress for nearly a century ; but tliis argument 
cannot be pressed, since such a question might be attected 
by the idiosyncrasies of an individual book-lover. The 
safest conclusion appears at present to be that the 
probabilities point to the second century as the date 
for this handsome MS., but that it would not be safe 
to treat it as an ascertained fact which may be used for 
the dating of other papyri. There can be very little doubt 
that the point will be settled before long by the discovery 
of fresh materials ^. 

A short mention is perhaps due, in conclusion, to a MS. The 
of considerable literary importance belonging to this omtion of 
period, though it falls rather outside the sequence of ^^P*'''" 
palaeographical development; namely the papyrus which 
contains the Funeral Oration of Hyperidesl Like the 
papyrus of Aristotle, the oration is written on the verso 
of the roll, and in a private and non-literary hand. It 
is not even, as the Aristotle is, a copy made by some 

' An important piece of evidence in support of the conchision stated 
above has already come to light in one of the Oxyrhynchus papyri (Gren- 
fell and Hunt, Oxijrhynchus Papyri, i no. 20, plate v). This is, curiously 
enough, also a copy of the second book of the Iliad, and is w^ritten in 
a large and ornamental hand, strongly resembling that of the Oxford 
Homer ; while its date is approximately fixed by the occurrence on the verso 
of some accounts which cannot be later than the third century. The 
Oxford Homer may, therefore, now be assigned to the second century with 
considerable confidence. 

^ Brit. Mus. Pap. xcvin ; complete facsimile (by hand) in the cditio 
princeps by Churchill Babington ; specimen photographic facsimile in 
Cat. o/Anc. MSS. {Greek), plate iv. 


private person desirous of possessing a work of classical 
antiquity. The character of the handwriting and the 
multitude of mistakes show that it must be a mere school- 
boy's exercise. Its literary value, in spite of the many 
blunders of transcription, is very great, since it alone 
preserves for us one of the most famous speeches of an 
orator who stood among the first in a time when the 
oratory of the world reached its highest pitch; but for 
the humbler purposes of palaeographj^ it is not very useful. 
A schoolboy's exercise can throw little light on the writing 
of trained literary scribes, and holds no place in the 
history of palaeography. The date of the MS. is given 
approximately, though not preciselj^, by the writing on 
the recto, which consists of a horoscope, the year of which 
is shown by astronomical calculations to be either a. d. 95 
or A. D. 1 ^^, with a decided probability in favour of the 
former. A horoscope would probably lose its interest 
after the lapse of a generation or so, and the oration 
may consequently be dated with practical certainty in 
the course of the second century^. The horoscope itself 
is written in a neat and graceful hand, which approaches 
the literary type, and serves to confirm the description 
which has already been given of the Roman hand. The 
same may be said of another horoscope in the British 
Museum (Pap. ex), which is dated in the year 138 ^. 

The third The character of the book-hand of the first and second 

century, centuries has now been established as fully as the extant 

materials allow, and all the important documents which 

' A MS. which in some degree recalls the writing of the Funeral 
Oration is published by Messrs. Grenfell and Hunt among the Oxyrhynchus 
papyri (no. 9, plate iii). It is a fragment of a treatise upon metre, 
probably by Aristoxenus, written in a rather ugly hand of medium size, 
sloping backwards, and in dark ink. It may be assigned to the latter 
part of the second or to the third century. 

^ Facsimiles of both these horoscopes are given in the atlas accompanying 
vol. i. of Greek Papyri in the British Museum. 


come within that period have been described. In passing 
to the third century we enter a region much more obscure. 
There are no MSS. which can be assip-ned to it with 
absolute certainty, or even with such a clear approach 
to certainty as was the case with many of the papyri 
which have hitherto been discussed ; and there are not 
many which can be placed in it on reasonably probable 
evidence. There is, however, a group of MSS., exhibiting 
the same general type of hand, which, in the light of 
recent discoveries, appears to belong to this century; and 
it is a group of some importance in palaeographical histor}^ 
The evidence on which this date is assigned to it is cir- 
cumstantial rather than direct; but it seems to be fairly 

The MS. on which the decision of the question mainly Brit. Mus. 
turns is a papyrus in the British Museum containing the ^^P-^^^"^^- 
second, third, and fourth books of the Iliad^. Unlike 
all the papyri which have hitherto been described, it is 
in book form, consisting of eighteen leaves, the text being 
written on one side only. The most noticeable character- 
istic of the hand in which it is written is that it has 
a decided slope. This in itself marks it off from the 
upright hands of the first and second centuries, and it 
also produces a lateral elongation of many of the letters 
which is quite alien to the square build of the earlier 
characters. A becomes broad, with a tendency to revert 
to the angular shape of the later Ptolemaic style ; M shows 
a similar tendency, becoming at once wider and shallower 
than before ; € is less well rounded ; is very small ; Y is 
angular and Y-shaped. The general aspect of the hand 
is rough and coarse, but this may be due to the scribe, 
who certainly was not a well-educated man, rather than 
to the times in which he lived. 

' Pap. cxxvi ; specimen facsimile in Classical Texts, jjlate vi. The cata- 
logue of ships is omitted from Book ii. and only the first hundred lines of 
Book iv. are preserved. Alphabet in App. I, col. 18. 


There is no direct evidence of the date of the MS., but 
some indirect evidence is provided by the writing which 
stands on the verno of the last four leaves, containing 
a short grammatical treatise attributed to Tryphon ^ It 
is written in a fairly good cursive hand, rather small 
and compact, and leaning somewhat to the right. Until 
quite recently there was nothing to fix the date of this 
hand, so that it was of no use in establishing the age of 
the writing on the recto ; but the clue has been fur- 
nished by some of the papyri brought home in 1896 by 
Mr. Grenfell. Some of these are dated in the reign of 
Diocletian, in the years just on either side of a. d. 300 ; 
and a comparison shows great similarity between these 
liands and that of the Tryphon. If, then, the Tryphon is 
to be placed somewhere about the year 300, the Homer 
must have been written earlier, and may safely be as- 
signed to the third century. These dates are earlier 
than those which have previously been attached to the 
MS. ^ ; but whereas the former opinion rested upon little 
but conjecture, the new dates are based upon fairly good 

The results of this discovery are of some importance. 
In the first place it seems to carry with it the dates of 
a certain number of other documents, to which no dates 
could be satisfactorily assigned before, but which exhibit 
a sloping hand akin to that which has been described. 
For instance, there is a fragment of the well-known 
The Berlin theological work, the ^liepherd of Hermas, published 
Hernias. ^^^j^Q^gj^ ^^^^ identified) by Wilcken^ The editor of this 
fragment, who stands quite at the head of the continental 
students of papyri, could not, at the time of the publica- 
tion (J890), make any closer approximation to its date 

' Specimen facsimile in Classical Texts, plate ix. 

- In Classical Texts the Homer is provisionally assigned to the fourtli 
or fifth century, the Tryphon to the fifth or perhaps the sixth. 
" Tafeln zur dlferen griechischen Paldocjraphie, plate iii. 


than to say that it must be some centuries older than 

a certain vellum MS. of the eighth century (the Frag- 

mentuni Mathematicum Bobiense) to which it bears some 

general resemblance. It is now, however, fairly clear 

that it must be placed in or near the third century, 

being written in a sloping hand akin to that of the British 

Museum Homer, though somewhat finer and thinner in 

style. The letters are of much the same shape, except 

that the curve of M is rather less shallow, and the hand 

as a whole shows the same evidence of looseness and 


Another frao-ment of some interest is the small piece Other 

. MSS. in 

of Ezekiel discovered by Mr. Grenfell, and now in the sloping 

Bodleian, containing a Hexaplar text, with a few of the li^nd. 
critical symbols employed by Origen\ Mr. Grenfell's 
second volume contains three additional examples of this 
type of hand, which may all be assigned to the same 
century ; namely, the very interesting scrap of Pherecydes, 
a portion of a drama (there identified with the Melanippe 
Desmotis of Euripides, but perhaps a comedy), and parts 
of two columns of a prose work, which has been identi- 
fied by Prof. Blass and others as a portion of the Memora- 
bilia of Xenophon^. In the case of the drama, the 
evidence of date is strengthened by its possessing some 
scholia in a cursive hand which cannot be later than the 
third century. 

The existence may consequently be taken to be estab- 
lished of a sloping type of hand, which is characteristic 
of the third century; and this confirms the conclusion 
to which reference has already been made, that the history 
of writing upon papyrus anticipates that of writing upon 
vellum, so that similar phenomena on difterent materials 
do not indicate contemporaneousness in date, but, on the 

' Grenfell, Greek Pcqw^, J- 5 (without facsimile). 

■^ Grenfell and Hunt, Greek Papyri, ii. 11-13 ; facs. of the Pherecydes and 
the drama, in plates iv and iii respectively. 


contrary, admit of a difference of some centuries. On 
vellum a sloping hand is first found about the eighth cen- 
tury ; but no one would propose to place the papyri which 
exhibit a sloping hand so late as this, when Greek writing 
was practically extinct in Egypt. If, however, these 
papyri are put back to the fifth or sixth century, as has 
hitherto been done, the principle of conformity with the 
vellum MSS. is abandoned, and there is consequently no 
reason why we should not carry them yet further back 
to the third century, if, as is now the case, trustworthy 
evidence points to such a conclusion. 
The It, so happens that papyri of this type are, at present, 

Marseilles hi • c ^ • 

isocrates. not at all plentiful m number, and nearly all that exist 
are mere scraps. Besides the Homer MS. already described, 
only one literary document of any size can as yet be 
placed in this category, namely the Marseilles papyrus 
of Isocrates In Nlcoclem ^. Concernino; the date of this 
MS., which is of some value to the textual tradition 
of Isocrates, very diverse opinions have been expressed ; 
the first editor, Schoene, assigning it to the Ptolemaic 
age, while B. Keil, who followed Schoene in discussing 
the MS., places it in the fifth century after Christ. The 
former date is manifestly impossible ; the latter is probably 
too late. The MS. is written in two hands, the first 
a light, sloping hand, with affinities to those which have 
just been described, while the second is solider and heavier, 
having no very close parallels among extant papyri. It 
is, of course, possible that this hand shows us the successor 
of the third-century hand, but the differences in the shapes 
of individual letters are not very great, and it is more 
likely that both hands belong to the third century, and 
only differ through the personal idiosyncrasies of different 
scribes '^. 

' Two specimen facsimiles (showing the diiferent hands) in Melanges 
Graux, with article by E. Schoene. 

^ The Oxyrhynchus papyri contribute some additional examples of this 



In addition to this sloping book-hand, a semi-literary 
hand of a very different appearance is found in use in 
the latter part of the century. It resembles the cursive 
hand of the same period, which has been described in 
chapter III (p. 46), and is, in fact, that cursive hand 
adapted for literary use. An example of it may be seen Britisii 
in a fragment of a Psalter in the British Museum \ psaiter 
which may apparently stand with the Oxford Ezekiel, ^^P- 
a page of St. Matthew among the Oxyrhynchus papyri, 
and a scrap of Isaiah in the Rainer collection at Vienna, 
as the earliest extant remains of the Greek Bible. The 
hand is upright and somewhat showy, with the letters 
K and especially prominent ; but as it does not appear 
to be a regularly-established book-hand, it need not be 
described at length. The Psalter in question was appar- 
ently used as a school-book, having the separate syllables 
marked off by a different hand, to which is probably also 
to be ascribed some writing- on the verso in which the 
syllables are actually written apart. This writing on the 
verso is in a cursive hand which belono-s to the third or 
fourth century, and consequently confirms the date which 
has been assigned to the Psalter. 

A hand very similar to that of the Psalter, though 
rather smaller, is found in a long magical papyrus in 

sloping third century hand. The most important of these from the 
literary point of view is the fragment which is attributed, with great 
probability, to Sappho (no. 7, plate ii), and which is palaeographically 
notable as containing the only written example of a digamma, with the 
exception of one in the Paris Alcman. Still more imi^ortant palaeo- 
graphically is a fragment of Plato's Laivs (no. 23, plate vi), on the verso 
of which is a date in the year A. d. 295, and which consequently confirms 
the attribution of hands of this type to the third century at latest. A 
fragment of St. Matthew (no. 2, plate i), which is likewise assigned to the 
third century, is written in too rough and untrained a hand to be of 
much palaeographical value. 

^ Brit. Mus. Pap. ccxxx ; a roughly printed facsimile in A/Iiencmm for 
Sept. 8 1894. The arguments there adduced for believing tlie papyrus 
to have been in book form do not appear to be sound. 


the British Museum ^ ; the semi-literary character of the 
contents being suited well enough by the semi-literary 
style of the writing. 

The consideration of these MSS. has brought us to the 
end of the third century, and here, with the re-organization 
of the Roman world by Diocletian, the Roman period of 
palaeography comes to an end too. In the domain of non- 
literary papyri we have seen that this change of government 
is accompanied by a change in the current style of writing, 
which makes the division into periods real, and not merely 
arbitrary. In the department of literary palaeography 
there is likewise a change, which comes to pass at about 
the same date, but it is a change of a different character, 
the nature of which will be considered in the following 

To sum up in brief the history of the period which has 
just been described, it will be seen that it covers a space 
of some 350 years, as against the 250 or 200 which are 
all that the present limits of our knowledge give to the 
Ptolemaic asfe. The main characteristic introduced into 
Greek writing during this period is that of roundness, 
which is accompanied by increased fluency of style. 
Letters are formed with curves instead of angles, and they 
are written, as much as possible, without raising the pen. 
The half-century which lies before the beginning of the 
Christian era concludes the period of transition which had 
begun under the later Ptolemies, and by about a. d. i the 
Roman style, with its rounded curves and square build 
of letters, is well established. The first century is the 
period of its perfection ; but before the close of it signs 
of decay appear in the greater slackness and looseness 
of formation in many letters. During the second century 
this tendency is increased, and the appearance of firmness 
combined with grace, which marks the earlier Roman 

' Brit. Mus. Pap. cxxi ; facs. in atlas accompanying vol. i. of Greek Papyri 
in the British Museum. 


hand, disappears from all except the best MSS. In the 
third century uprightness is exchanged for a sloping hand, 
— always a sign of degeneracy in palaeography ; and this 
is generally accompanied hy looseness and irregularity 
in the individual letters. The Roman hand has plainly 
run its course, and there is room for the new departure 
which has now to Ije considered. 



No Byzan- In treating of the non-literary papjn-i it has been shown 
o fTiteriiry ^^^^ ^^® Roman period is followed by an equally long and 
papyri. prolific period, to which the name of Byzantine has been 
given. But in the sphere of literary papyri there is no 
Bj^zantine period. It is true that a certain number of 
literary texts upon papyrus have been, at various times, 
located between the fourth and the seventh centuries ; 
but reason has been shown in the previous chapter for 
transferring some of these to much earlier dates, while 
those which remain are, as will be seen presently, few 
and unimportant. Indeed the student who is acquainted 
with the principal extant papyri will already have re- 
cognized that hardly any of them still remain to be 
dealt with. The explanation is not to be found in any 
scarcity of papyrus as a writing material. The masses 
of documents of the sixth and seventh centuries which 
have been found in Egypt are sufficient to disprove any 
such theor}^. The cause is different, and is to be found 
in the successful rivalry of another material, namely 
parchment or vellum. 
Early use The beginnings of the use of vellum are to be found 
o ve um. j^^g]-^ earlier, and are recorded in the standard works 
on palaeography. Without discussing the precise value 
of the story which tells of its invention in Pergamum 



in the second century before Christ, or the reference by 
Cicero to the use of hif^dipai'^, it is clear from Martial's 
Ax>ophoreta that vellum was considerably used as a writing 
material in the second century of our era. It seems 
certain, however, that at this time it was generally 
regarded as inferior to papyrus. It was used for note- 
books, or for rough copies of literary compositions; and 
if, as Martial shows, copies of classical authors, such as 
Homer, Virgil, Livy, and Ovid, were sometimes made on 
vellum, they were common and cheap reproductions, better 
to give away to clients and hangers-on than to keep as 
ornaments of a good library. 

The earliest extant example of writing upon vellum Fragment 
is perhaps a specimen of this very class of book. It is sthenes 
a somewhat recent acquisition of the British Museum, vellum, 
consistinof of two small leaves of vellum, with two columns 
to each page, containing part of Demosthenes' speech De 
Falsa Legatiome^. This fragment was found in Egypt, 
and though the exact circumstances of its discovery are 
not known, it came to Europe among a large collection 
of papyri, the majority of which belonged to the Eoman 
period. The hand in which it is written bears a decided 
resemblance to that of the Herodas papyrus. It is quite 
unlike any known vellum hand, and we are therefore 
justified in regarding it as belonging to the period to 
which its appearance seems to assign it, namely, about 
the second century. It may thus fall within the period 
at which Martial was writing, and it corresponds very 
well to what has been said of the character of the 
vellum MSS. mentioned by the poet. It is plainly not 
an elaborately written copy. There is nothing of the 
appearance of an edition de luxe, as in the Vatican or 
Sinaitic MSS. of the Bible, or the Palatine and Medicean 

' Epp. ad Att. xiii. 24. See Thompson, pp. 35-37, for a summary of the 
data on this subject. 

^ Brit. Mus. Add. MS. 34473 ; text and description in Journal 0/ Philology, 
xxii. 247 (1894). 



Virgils. It may well have been regarded as an inferior 
class of book to the best papyrus MSS. of the period. 
Absence of For the present, this example stands alone, and there 
mnvH^ is every reason to suppose that to the end of the third 
after third century papyrus held its own, at any rate in Egypt, as 
^'. the material on which literary works were written. But 
when we come to the fourth century the situation is 
. - changed. For non-literary purposes, as has been shown 
^ above, the Roman period is succeeded by a Byzantine 

•- period, equally long and equally prolific, lasting well 

into the seventh century. Literary palaeograjDhy, how- 
ever, has no Byzantine period, so far as papyri are con- 
cerned. It cannot be due to accident that scarcely any 
literary works on papyrus have come down to us which 
can reasonably be assigned to dates between the fourth 
and seventh centuries. It is true that there have been 
periods which, until quite lately, were unrepresented 
among extant papyri ; but the cases are not parallel. 
In those instances no papyri at all of the periods in 
question had been discovered ; which merely meant that 
the spade of the explorer had not hit upon any site of 
that particular date. In the case of the Byzantine period, 
' on the contrary, not hundreds merely, but thousands 
of papyri have been found, which now choke the libraries 
of Berlin and Vienna; yet among these scarcely one 
contains literary work, and the few that do are palpably 
not written by professional scribes for the j^ublic market. 
It is, therefore, a fair conclusion that papyrus was not 
at that date being used so freely for literary purposes 
as it had been during the Roman period. 
Vellum This inference from the negative argument of the 

the fourth absence of literary papyri later than the third century 
century, jg confirmed by positive evidence from the other side. 
The fourth century is the date to which our earliest 
extant vellum MSS. (the above-mentioned Demosthenes 
excepted) are assigned ; and though the grounds for this 


ascription are not quite so conclusive as one could wish, 
they are sufficient to justify a fair amount of confidence. 
The MSS. in question include the Vatican and Sinai tic 
Codices in Greek, the Codex Vercellensis of the Old Latin 
Bible, the Codices Palatinus, Romanus, Vaticanus, and 
Sangallensis of Virgil, and possibly the Bembine Terence. 
The period of transition is further marked by the well- 
known statement of Jerome, that Acacius and Euzoius 
(towards the end of the fourth century) caused a number 
of damaged volumes in the famous library of Pamphilus 
at Caesarea to be replaced by copies upon vellum. The 
language employed by him ^ implies that this was a 
deliberate and systematic attempt to renovate the library ; 
and the adoption of vellum in place of papyrus indi- 
cates that the victory of the former material had been 
gained by this time. It is consequently certain that the 
introduction of vellum had begun at a considerably earlier 
date. Finally, the evidence is clinched by the statement 
of Eusebius that the emperor Constantine, after his 
acceptance of Christianity, ordered fifty copies of the 
Scriptures to be prepared on vellum for use in the churches 
of Constantinople. When an emperor, on so special an 
occasion, orders vellum copies for the churches of his 
capital, in the beautifying of which no expense had been 
spared, we may be confident that vellum was by that time 
regarded as the best material for literary purposes. 

It does not follow that papyrus was at once and alto- A few 
gether extinguished by the victory of vellum, and there MSS.^of 
remain a few MSS. to be mentioned which appear to ^o^^tli 
belong to the fourth or later centuries. In no case, how- or later, 
ever, are they documents of much palaeographical im- 
portance. A very long magical papyrus at Paris ^, and 

* 'Quam [bibliothecam] ex parte corruptam .... in membranis 
instaurare eonati sunt' {Ep. 141). 

^ Specimen facs. \nOn\oni's Facsimiles des;plus anciensManuscritsGrecs. .. 
de la Bibliotheque Rationale, plate i. 

I 2 


a less long one in the British Museum \ have been 
assigned to the fourth century, though there is no direct 
evidence to fix their date. They are written in book 
form, on rather narrow pages (as in the case of the 
papyrus of Iliad ii-iv, which has been assigned above 
to the third century), and the writing slopes somewhat; 
but it is not quite of the same type as the sloping hands 
of the third century. It is a rounder, thicker, harder 
hand, and its appearance suggests the possibility of the 
writer having^ been familiar with vellum MSS. Further, 
the text is written on both sides of the leaf, unlike the 
Homer papyrus mentioned above ; and this again seems 
to indicate the influence of vellum. On the whole, the 
case is one on which the only safe plea is that of ignorance ; 
but such indications as there are seem to point to the 
fourth century rather than the third. The contents of 
both documents are strongly coloured by Gnosticism, but 
so intermixed with other elements that it is not at all 
likely that they were written during the flourishing time 
of that philosophy ; while mistakes are so frequent as 
to show that they are somewhat unintelligent copies, and 
not original works. So far, therefore, as the contents 
are concerned, there is nothing to discredit the date that 
has generally been assigned to these MSS. 
Psalter Of more interest is a papyrus Psalter in the British 

Museum. Museum ^, as to the age of which exceedingly different views 
have been held. It consists of thirty-two leaves, rather 
short and square in shape, containing Psalms x-xxxiv, 
written in an irregular sloping hand. Tischendorf, who 
first published its text, regarded it as earlier than any 
vellum MS. in existence ; but there is no reason to suppose 
that he had ever given much attention to the palaeography 

' Pap, XLVi ; facs. in Atlas accompanying vol. i. of Greek Papyri in the 
British Museum^ plates xxxvi — xlix. 

^ Pap. XXXVII ; specimen facs. in Cat. ofAnc. MSS. {Greek), plate xii ; Pal. 
Soc. i. 38. 


of papyri, nor was there then sufficient evidence extant 
to enal)le him to form an opinion. The editors of the 
Palaeographical Society at first left it as early as the 
fourth or fifth century, but subsequently brought it 
down to the sixth or seventh; and there is little doubt 
that the later date is correct. The hand is not that of 
a trained scribe at all, and is not of the literary type. 
It is not in the least like the sloping hand which we 
have assigned to the third century. It is closely akin 
to the non-literary hands of the late Byzantine period, 
described in Chapter III. It will be remembered that 
a sloping hand comes into use for non-literary documents 
towards the end of the sixth century, and continues 
into the seventh; and it is to this class of hands that 
the Psalter in question belongs. It is of some importance 
to fix the date, on account of the interest that would 
attach to the MS. if Tischendorf's claim could be made 
good ; but its palaeographical importance is not great. 

So far, the late papyrus hands have had no direct Bntisii 
and close connexion with the contemporary vellum hands. Hesiod. 
There are, however, some cases in which there is a con- 
siderable resemblance. The earliest, in all probability, of 
such MSS. is a fragment of Hesiod in the British 
Museum ^ written in a large uncial hand of the same 
general type as the Codex Alexandrinus. It is an up- 
right hand, square in formation and well-rounded, but 
differing from the second century MSS., of which the 
same description has been given, in the thickness and 
coarseness of the writing. This does not appear to be 
due to a lack of skill on the part of the scribe, but to 
be the result of deliberate intention. It is rather an 
imitation of the earlier hand at a time when the light- 
ness of touch, so necessary for good writing upon papyrus, 
had been lost; and consequently there seems to be no 

' Pap. CLix ; text in Revue de Philologie, xvi. i8i. 



reason for separating it from the vellum hands to which 
it appears, prima facie, to be akin. 
The Another papyrus, reproducing even more evidently 

St*! Cyril, the writing of contemporary vellum MSS., is a MS. at 
Dublin which contains some portions of the treatise De 
Adoratione of St. CyriP. It is written in page-form 
in a good-sized uncial hand of the type that is generally 
called Coptic. It is especially marked by the upright 
A and the deeply curved M (M), which are found in 
a few vellum MSS., such as the Codex Zacynthius and 
the Codex Marchalianus. These MSS. are generally 
assigned to the sixth or seventh century, and the papyrus 
of Cyril may reasonably be placed in the same period. 
In a very similar hand is the fragment of a Festal Letter 
from a patriarch of Alexandria to his clergy, w^hich was 
' obtained by Mr. Grenfell in 1895-6, and of which the 
date is probably A. D. 577 ^. 
Papyrus of The last MS. that need be mentioned in this connexion 
Prophe'tr i^ one of the prophets Zechariah and Malachi, which 
was formerly (and probably is still) in the possession 
of a Viennese dealer, Theodor Graf. It was exhibited 
at the London Oriental Congress of 1893, with the state- 
ment that its date was in the second century^. This 
date is, however, wholly impossible. The MS. is really 
of the same type as the last, but larger, rougher, and 
more irregular. It is a large, coarse uncial hand, written 
on pages of a good size. The leaves are bound together 
into quires by pieces of string. The seventh century is 
probably quite as early a date as it is entitled to claim, 
and, like the Cyril MS., it falls rather outside the domain 
of papyrus palaeography. The material on which it is 
written is indeed papyrus, but the writing is that which 

* Published with facsimile in Transactions of tJie Royal Irish Academy, vol. 
xxix, pt. xviii, by the Kev. J. H. Bernard. 

2 Published with facsimile in Grenfell and Hunt's Greek Papyri, vol. 
ii ; now Pap. dcoxxix in the British Museum. 

^ A rough facsimile was given in the Times during the Congress. 


belongs properly to vellum. It shows a late stage in the 
transition from papyrus to vellum, when the victory of the 
latter had been won, though the Coptic Church continued 
for some centuries to use the ancient material for its ver- 
nacular Bible, and also, to some extent, for Greek copies 
of the Scriptures also. 

Conversely, there are one or two MSS. which show upon Veilum 
vellum a style of writing to which we are accustomed onlv ^^^' ^" 

, " papyrus 

in papyri. The first of these, the fragment of Demosthenes' styif. 
De Falsa Legatione, has been described already : the 
other which seems to deserve notice is the MS. of the 
Gospel and Revelation of Peter, discovered at Akhmim 
in 1886 and published in 1893^. This has been assigned 
to the eighth century by some palaeographers, including 
the very high authority of M. Omont ; but in truth there 
is no even approximate parallel to it among vellum MSS. 
Its kinsmen must be found in the Byzantine papyri of the 
sixth and seventh centuries, especially the former, as they 
have been described in Chapter III; and there seems to 
be no reason for declining to give this date to the Peter MS. 
It is confirmed by the other contents of the same volume, 
which are written in hands more closely conforming to 
the vellum type, and apparently of the sixth century. 

Instances such as these, however, whether of vellum The 
MSS. written in papyrus hands, or of papyrus MSS. [^''"ejlu'" 
in hands which belong properly to vellum, are so scarce 
as to be plainly exceptional, the natural concomitants 
of a period of transition. The general conclusion seems 
certain, that about the fourth century papyrus ceased to be 
the normal material for works of literature, and its 
place was taken by vellum. Of course this change of 
material involved a certain amount of change in the style of 
writing ; but the general type remained the same, or rather, 
developed only in an ordinary manner and at an ordinary 

' Full facsimile by the French Mission Arch^ologique at Cairo ; also by 



.^ rate. The earliest vellum MSS, carry on the traditions 
I of the Roman style of writing upon papyrus. The only 
change that is apparent is one in the direction of greater 
. -v. care and precision. The scribes found that the new 
'" , " material lent itself very easily to a handsome style of 
writing ; and they consequently abandoned the somewhat 
slovenly sloping style which had come in during the third 
century, and modelled themselves on the earlier and 
more careful writing of the first and second centuries. 
The resemblance of the Sinaitic and Vatican codices 
• to papyrus MSS. has been noticed more than once. Not 

only in the narrow columns, arranged four or three on 
a page, and showing eight or six at each opening of the 
book, but also in the hands themselves, there is a very 
real resemblance to .papyri of the best Roman period. 
The Vatican MS. recalls such a papyrus as the British 
Museum Odyssey, or that which contains books xiii, xiv 
of the Iliad; the Sinaitic MS. reminds one still more 
strongly of the great Hyperides papyrus. The chief varia- 
tion is in the increased thickness of the heavier strokes, 
which are differentiated from the light strokes far more 
than was natural on the comparatively delicate papyrus ; 
though this tendency actually reacts upon the pajoyri 
written after the vellum period had set in, and accounts 
for the greater heaviness and thickness mentioned above 
as characteristic of such papyri. 
The dates The question is sometimes asked whether the evidence 
early*' derivable from papyri does not modify our views as 
vellum to the dates of the early vellum uncials. It has been 

uncials. . t i 

shown m the preceding chapter that a hand or the same 
general type as that of the vellum MSS. is found on 
papyrus as early as the end of the first century. Is it 
not possible that the great MSS. of the Greek Bible are 
earlier than they have hitherto been supposed to be ? 
It is impossible to argue the question fully here, since 
its decision rests largely upon details not properly palaeo- 


graphical, such as the presence or absence of certain 
divisions of the text of the New Testament (the Eusebian 
canons, the Ammonian sections, the Euthalian divisions of 
the Acts) ; but it may be stated that the palaeographical 
evidence does not require any departure from the dates 
which have become traditional. In the case of the Codex 
Alexandrinus, the presence of a treatise of Athanasius 
(attached to the Psalms) precludes a date earlier than 
the latter part of the fourth century; while the Codex 
Sinaiticus has two notes or colophons which contain 
references to the library of Pamphilus at Caesarea — a 
library only formed at the beginning of the fourth 
century. On the other hand, the style of the beginning 
of the sixth century is fixed by the Vienna MS. of Dios- 
corides written for Juliana Anicia, daughter of Flavins 
Anicius Olybrius, who was emperor in the west in a. D. 472. 
A fuller discussion of this subject would belong rather 
to an essay on vellum-palaeography ; but the facts here 
summarized show that the dates of the early Bible uncials 
do not rest upon palaeographical considerations alone, 
and are too firmly fixed to be aflected by the evidence 
of the papyri. The same cannot be said of the very 
fine MS. of Homer, with illustrations, now at Milan '. 
The hand of this has a particularly early appearance ; and 
though it would be rash to express a definite judgement 
merely on the strength of a few facsimiles, it is worth 
while suggesting a doubt whether this MS. may not be 
considerably older than the fifth century, the date to which 
it is now assigned. 

The conclusion, therefore, to which the evidence leads 
is that the literary tradition passed wholly from papyrus 
to vellum in the fourth century, the exceptions in either 
direction being too few and too unimportant to require 
much notice. It appears, further, that the hands of the 
early vellum uncials are the direct descendants of the 

^ Specimen facsimiles in Pal. Soc. i. 39, 40, 50, 51. 


hands found in literary papyri, especially those of the fii'st 

and second centuries. The results of this revolution are 

not solely of palaeographical importance. They have also 

had a bearing on literary history, and have fundamentally 

affected our present knowledge of ancient literature. 

Results of So long as papyrus was the material, and the roll the 

frompapy- form, in common use for literary purposes, it was impossible 

rus to ^Q include works of very great extent in a sinp-le volume. 

vellum. 1 

Very large rolls were cumbersome to use, and perhaps 
were liable to be torn more easily than those which 
were lighter and shorter ; in any case, as has been shown 
in an earlier chapter, a length of 20 to 30 feet is the 
greatest to which a single roll attains, while most of those 
which are now extant must have been considerably shorter. 
Consequently works of any great length could not be 
contained in a single roll, but had to be divided into 
a number of rolls. No papyrus roll of Homer hitherto 
discovered contains more than two books of the Iliad ^. 
Three short orations fill the longest extant roll of Hy- 
perides. The 'MTqvaiodv floAireta occupies no less than four 
rolls, though these, it is true, are of very small dimensions. 
A single play, a single long oration, would be enough 
to fill an ordinary roll. In the domain of sacred literature, 
the four Gospels and the Acts must each have occupied 
a separate roll ; and no complete New Testament was 
possible until vellum came into use. It is obvious that 
conditions such as these favour selection among the works 
of an author. The ordinary literary student would not 
possess a complete Aeschylus or Sophocles, but only his 
favourite plays ; and, since opinion tends to run in similar 
directions, some plays would be very much more in request 
than others. It is this fact that accounts for the state 

' There is perhaps an exception to this rule in a papyrus recently 
discovered by Mr. Grenfell, and now in the Bodleian, the fragments 
of which contain lines from books xxii, xxiii, and xxiv. This MS. is in 
the small hand of the third century b. c, which would admit of a great 
amount of text in a roll of ordinary size. 


in which the tragedians have come down to us. We 
have only a small proiDortion of their plaj^s, but we have 
all that were most admired in antiquity. The meaning 
of this is that, at the time when vellum came into use 
and large collections became possible, the great majority 
of the dramas had already become scarce, and consequently 
few save those which we now possess were transcribed 
into vellum codices. We can imagine a similar state of 
things if the plays of Shakespeare had continued for two 
or three centuries, or even more, to circulate only in their 
original quarto form, where each play occupies a single 
volume, instead of being collected into the folios shortly 
after his death. In that case posterity would, no doubt, 
possess Hamlet, Macbeth, As You Like It, and Much Ado 
about Nothing; but it is highly probable that Timon 
of Athens, All's Well that Ends Well, Pericles, and Titus 
Andronicus would have been lost, or would survive only 
in fragments. 

Other causes no doubt co-operated largely in determining 
the survival or disappearance of the various works of 
ancient literature. The stream of time, though far from 
acting as Bacon, in a pessimistic frame of mind, declared, 
has yet been singularly capricious in its choice of the 
cargoes which it has brought safely down to us. But 
among the conditions which have exercised an important 
influence upon its operations are, beyond doubt, the use 
of papyrus and of the roll-form as the vehicle of literary 
tradition for six centuries after the great Attic period, 
and its supersession by vellum, and by the modern book- 
shape or codex, in the fourth century of our era. 

The influence of papyrus palaeography does not, however, The 
end here. For non-literary purposes, for the ordinary ^^^^^^ ,,^ 
occasions of daily life, papyrus continued to be used vellum 

. developed 

almost exclusively, at any rate in Egypt, and apparently from late 
elsewhere, for several centuries after its practical dis- P'^Py/^^^ 

'- ^ hands. 

appearance from the literary field. The develoj^ment of 


cursive writing for private and business purposes runs 
its regular course, as has been traced in the third cliapter 
ot" this essay, until the seventh, or even the eighth century. 
Then our evidence ceases, through the accidental fact that 
the use of Greek was about that time obliterated in the 
only country from which materials for this dej)artment 
of palaeography are derivable. But common sense tells 
us that elsewhere in the Greek-speaking world — in Con- 
stantinople, in Athens, in Asia Minor — Greek cursive 
writing continued to be used for every-day jDurposes, 
whether upon papyrus or (as no doubt was increasingly 
the case after the severance of Egypt from the Empire) 
on vellum. The stream runs underground for a space, 
but it reappears at the time of the palaeographic revolution 
of the ninth century, when minuscules superseded uncials 
as the current book-hand. As has been briefly stated by 
Wilcken ^, in a sort of obiter dictwin, the minuscule of the 
tenth century is the descendant of the papyrus-cursive 
of the seventh. The interval of time is so considerable 
that very close resemblances are hardly to be expected, 
especially since the tenth-century hand was a refined 
and idealized reproduction of its archetyjje; but the 
resemblance is quite traceable. The forms of letters in 
the late Byzantine papyri are already approaching those 
of the vellum minuscule ; and when they are written 
small, as in the receipts and accounts of the seventh 
centur}^, the resemblance is still more apparent. 77 has 
developed its h-shape, and fi has its long perpendicular 
tail ; y has its familiar v-shape, and v has lost its tail 
and become a simple curve ; a, 8, e, and indeed most 
of the letters, approximate to the forms in which we 
know them now. In the better examples there is even 
something of the sharpness and precision which mark 
the early vellum minuscules. They are not, it is true, 
nearly so beautiful as the MSS. of the tenth century 

^ In the Introduction to his Tafeln. 


— the perfection of Greek miniiscule writing — but they 
show plainly that they stand in the line of their ancestry. 

It is so that, for the first time, the minuscule hand 
enters the line of literary tradition. The hand which 
for fifteen hundred years had been confined, so far as 
its proper use was concerned, to accounts, wills, and other 
such legal and private documents, and had only acci- 
dentally, as it were, become at times the vehicle of 
literature, now, in an improved and purified form, becomes 
the regular repository of literary works. The minuscule 
hand, even in its most cursive manifestations, is as old 
as the set uncial, so far at least as our records go back, 
and comes down side by side with it, though confined 
to less dignified functions. But at the last, when the 
uncial hand was exhausted, and could no longer supply 
the needs of increasing literary demands, the minuscule 
hand came forward to take its place, and to carry on the 
torch of literature for the five hundred years that still 
remained before the invention of the printing-press. 

So, up to the end, the influence of papyrus remains 
traceable. Its immediate literary function was over when 
its book-hand gave birth to the vellum uncials of the 
fourth century ; but its non-literary hand carried on 
a tradition which was to claim its place in literature 
six hundred years afterwards. Then its duties were at 
an end, and it is only in its remote descendants that 
it still survives, in the Greek types of to-day, which have 
been imitated or developed from the written hands of the 
Middle Ages. 



The following table gives the alphabets of the chief literary 
papyri which have been described in the preceding pages. 
They are arranged in what is believed to be their chronological 
order, though in the cases of MSS. of the first and second 
centuries this order must be regarded as only approximate. 
In the process of reproduction the size of the letters has been 
slightly reduced throughout. The following are the MSS. 
from which the alj^habets are taken. 

1. Literary fragments (Grenfell and Hunt, Gk. Pap. ii. i 

— Brit. Mus. Pap. dclxxxviii). Early third centuiy b. c. 

2. Petrie Pliaedo (Brit. Mus. Pap. cccclxxxviii). Third 

century b. c. 

3. Petrie Antiope (Brit. Mus. Pap. cccclxxxv). Third 

century b. c. 

4. Dialectical Fragment (Louvre Pap. 2). Before 160 b. c. 

5. Hyperides, In AtJienogenem (in the Louvre). Second 

century b. c. 

6. Bacchylides (Brit. Mus. Pap. dccxxxiii). First century 

b. c, 

7. Hyperides, In Philij)jpiclem (Brit. Mus. Pap. cxxxii). First 

century b. c. 

8. Demosthenes, Ep. in. (Brit. Mus. Pap. cxxxiii). First 

century b. c. 

9. Herculaneum Pap. 152. First century b. c. Other 

Herculaneum MSS. have similar alphabets. 
10. Homer, i7/af/ xxiii, xxiv. (Brit. Mus. Pap. cxxviii). Late 
first century b. c. 



11. Petition (Brit. Mus. Pap. cccliv). Circ. lo b. c. 

12. Homer, Odyssey Hi. (Brit. Mus. P;ij). cclxxi), Circ. a. d. i. 

13. Homer, Hiad xviii. (Harris Homer = Brit. Mus. Pap. 

cvii). First century. 

1 4. Hyperides, InDemosthenem, etc. {'Bv'it. Mus. Papp. cviii,cxv). 

Late first century. 

15. Homer, Iliad xiii, xiv. (Brit. Mus. Pap. dccxxxii). First 


16. Homei', Iliad xxiv. (Bankes Homer = Brit. Mus. Pap. 

cxiv). Second century. 

17. Homer, Iliad ii. (Bodleian MS. Gr. class, a. 1 (P)). 

Second centuiy. 

18. Homer, Uiad ii.-iv. (Brit. Mus. Pap. cxxvi). Third 


For the non-literary papyri, an excellent table of alj^habets 
will be found in Sir E. M. Thompson's HandhooJc to Greelc and 
Latin Palaeography. 

































































































































































































2 • 





























































The following list was originally compiled independently, 
))ut since its preparation a similar list has been published by 
P. Couvreur in the Bevue de FMlologie ^xx. 165), and atiother 
by C. Haeberlin in the CentraTblatt filr Bihliothelswesen {xiv. i, 
2or, 263, 337, 389, 473), giving fuller details of the literature 
relating to each MS. From these lists some references have 
been taken ; but the magical papyri (the principal publications 
of which are mentioned in App. Ill) and a few other non- 
literary works included by Haeberlin are not given here. 
The Oxyrhynchus papyri do not appear in either of the above- 
mentioned lists. In the case of papyri not in the British 
Museum, and of which no facsimile has been published, the 
dates given are those assigned by their editors. Where none is 
given, it will be understood that the editors have given none. 

The following abbreviated references are used in this list : 

Cat of Additions = Catalogue of Additions to the Department of MSS., 

British Museum (published every six years). 
Cat. of Anc. Greek ilfSS. = Catalogue of Ancient MSS. in the Briti.-h 

Museum ; part I, Greek. Edited by Thompson and Warner (1881). 
Class. Texts = Classical Texts from Papyri in the British Museum ; 

ed. Kenyon (1891). 

F. or J^M/(7e/=Fiihrer durch die Ausstellung der Papyrus Erzherzog 
Rainer ; Greek section edited by Wessely (1894). 

G. P. = Greek Papyri ; vol, i by Grenfell. (1896), vol. ii by Grenfell and 
Hunt (1897). 

Mitlh. or Mitth. Ersh. EaOT«- = MittheiIungen aus der Sammlung der 
Papyrus Erzherzog Rainer. 



Notices ef Extraits = Notices et Extraits des Manuscrits de la Bibliothfeque 

Imperiale et autres Bibliotheques, vol. xviii (1865). 
0. P. =Oxyrhynchus Papyri ; ed. Grenfell and Hunt (part i, 1898). 
P. P. =Petrie Papyri ; ed. Mahaffy, two vols. (1891 and iSgaX 
PJiil. ^Hse/(/. = Philologischer Anzeiger (separate publication of Philo- 

logus) '. 
Rev. de Phil. = Revue de Philologie. 
Rhein. Mms. = Eheinisches Museum. 

Tafeln = Tixfeln zur alteren griechischen Palaographie; ed. Wilcken (1891). 
Wien. Sfttf/.= Wiener Stiidien. 
Zeitschr.f. ag. Sprac^c = Zeitschrift fiir agyptische Sprache. 

Aesehines, In Ctesiphontem, §§ 178-186. Fifth century. 
Kainer Coll. Mentioned by Karabacek, Mitth. Erzh. Bainer, 
i. 51. 

Aeschylus, Carians{7); frag, (fifteen lines). Before b. c. 161. 
Didot Papyrus. Weil, Monuments Gives (1879), with facs. 

— 3Iyrmidons{?) ; frag, (eight lines). Before b.c. 161. Didot 
Pap. Ed. Weil, with facs. 

Alcidamas, MovaeLov; fragment. Third century b.c. Brit. 
Mus, Pap. ccccxcix. Mahaffy, P. P. i. 25, with facs. 

Alcman, frag, of a lyric poem ; portions of three columns. 
First century. Louvre Pap. 71. 'Eggev, Memoires cVHist. 
anc. et de PMlol. 1863, p. 159 ; facs. in atlas to Notices et 
Extraits, xviii., pi. 50. 

Alcman (?), seven hexameter lines, three being imperfect, in 
Aeolo-Doric dialect. First or second century. Grenfell 
and Hunt, 0. P. i. 8, with facs. 

ApoUonius, Homeric Lexicon ; fragment. Late first century. 
Bodl. MS. Gr. class, e. 44 (P). Facs. by Nicholson. 

ApoUonius Rhodius; fragment. Eainer Coll. Mentioned 
in Phil. Anzeig. xv. 650. 

' A number of literai-y fragments among the Rainer papyri are briefly 
mentioned in various short notices in the Philologischer Ameiger for 1884-6 ; 
but it has seemed best not to include these, except v?hen they are fairly 
explicit. It is possible that mistakes were made in the first identifications, 
and until fuller details are published these fragments cannot be considered 
available for practical purposes. 


Aratus, PJiacnomena ; two fragments, in book form. Berl. 
Pap. Blass, Zeitschr. f. ug. Sp)-aclic, 1880. 

Aristophanes, Birds 1057-1081, 1101-1127. Louvre Pap. 
Weil, Bev. de Phil. 1882. 

Aristotle : — 

'AOrjvaiMv IIoAiTeta. End of first century. Brit. Mus. Pajj. 
cxxxi verso. Ed. Kenyon, 1891, with facs. ; -subsequent 
editions by Ferrini, Kaibel and Wilamowitz, Herwerden 
and van Leeuwen, Blass, Sandys. 

'AOrjvaLUiv IIoAtTeta, two fragg., in book form. Fourth cen- 
tury (?). Berl. Pap. 5009. Blass, Bhein. Mus. 1880; 
Diels, Abh. d. Berl. AMd. 1885, with hand-made facs. 

^Post. Analytics, i. 71 b 19 — 72 a 38 ; in book form. Seventh 
^ century (?). Berl. Pap. 166. Landwehr, PJiilologus, xliv. 21 

Aristoxenus (?), pvO/juKo. aroLX'^la, portions of five columns. 
Third century. Grenfell and Hunt, 0. P. i. 9, with partial 

Bacchylides, OfZes ; incomplete. First century b. c. Brit. Mus. 
Pap. Dccxxxiii. Ed. Kenyon (1897), with facs. ; sub- 
sequent editions by Blass and Jurenka (1898). 

Basil, E^yp. v. 77 e, vi. 79 e, ccxeiii. 432 b, cl. 239 c, 
ii. 72 A. Fifth centuiy (?). Berl. Pap. Blass, Zeitschr. 
f. d(j. Sprache, 1880; Landwehr, Philologus, 1884 (with a 

Bible : — 

Psalms xii. 7 — xv. 4. Late third century. Brit. Mus. Pap. 
ccxxx. Athenceum, Sept. 8, 1894, with facs. 

— xi. 2 — xix. 6, xxi. 14 — xxxv. 6, in book form. Seventh 
century. Brit. Mus. Pap. xxxvii. Tischendorf, Mon. Sac. 
Ined., nov. coll. i. 217; specimen facs. in Cat. of Anc. 
areelc MSS. 

— xl. 16 — xli. 4, in book form. Berl. Pap. Blass, Zeitschr. 

f. dg. Sprache, 1881. 

K 2 


Bible {continued) : — 

Song of Solomon i. 6-9, in book form. Seventh or eighth 

century. Bodl. MS. Gr. Bibl. g. i (P). Grenfell, G. P. 

i. 7. 
Isaiah xxxviii. 3-5, 13-16. Third century. Eainer Pap. 

8024 (Fiihrer, 536). 

Ezekiel v. 12 — vi. 3, in book form. Late third century. 
Hexaplar symbols. Bodh MS. Gr. Bibh d. 4 (P). Grenfell, 

a. p. i. 5. 

Zechariali iv — Malachi ?, in book form. Seventh century (?). 
Graf papyrus. Hechler, Times, Sept. 7, 1892, with speci- 
men facs. 

Matthew i. 1-9, 12, 14-20. One leaf of a book. Third 
century. Grenfell and Hunt, 0. P. i. 2, with facsimile. 

— XV. 12-16, fragment, Greek and Coptic, in book form. 
Sixth century. Kainer Coll. Mentioned by Gregory in 
Tischendorfs Novum Testamentum Graece iii. 450 (1884). 

— xviii, fragment, in book form (?). Eainer Coll. Fourth or 
fifth century. Mentioned by Gregory,. i&. 

Ilarl- XY. 29-38, in book form (?\ Fourth century. Eainer 
Coll. Mentioned by Gregory, ih. 

Luke V. 30 — vi. 4, in book form (attached to MS. of Philo, 
vicl. infr.). Fourth century. Paris, Bibl. Nat. Scheil, Mem. 
de la 3Iission Arch. Franc^aise au Caire, tom. 9 (1893), with 

— vii. 36-43, X. 38-42, in book form. Sixth century. 
Eainer Pap. 8021 (F. 539). 

Johni. 29, in book form (?j. Seventh century. Eainer Coll. 
Mentioned by Gregory, tilii supra. 

I Corinthians i. 17-20, vi. 13-18, vii. 3, 4, 10-14, imperfect, 
in book form. Fifth century (?). At Kiew, Uspensky 
Coll. Gregory, op. cit. iii. 344. 

— i. 25-27, ii. 6-8, iii. 8-10, 20, in book form. Fifth cen- 
tury. At Sinai. Eendel Harris, Biblical fragments from 
Mt. Sinai (1890). 



Bible {contimied) : — 

Uncanonical Gospel ; veiy small frag., parts of seven lines. 
Third century (?). Eainer Coll. Bickell, Mitth. Ersh. 
Rainer, i. 52 (1887), with facs. 

Logia lesu ; one leaf of a book. Third century. Egypt 
Exploration Fund. Ed. Grenfell and Hunt (1897), with 
facs. ; also in 0. P. i. i. 

Chronological Treatise; fragment, portions of six columns, 
covering B.C. 355-315. Third century. Grenfell and 
Hunt, 0. P. i. 12. 

Cyril of Alexandria, Be Adoratione, p. 242 e — 250 d, with 

lacunas, 286 b, in book form. Sixth or seventh century. 

Dublin Pap. Bernard, Royal Irish Acad. xxix. pt. 18, with 

partial facs. 
Demeti-ius ; philosophical works among Herculaneum papyri. 

First century b. c. See Scott, Fragmenta Herculanensia. 

Demosthenes : —  

In Arisiocratem, lexicon to ; fragment. Fifth century (?). 
Berl. Pap. Blass, Hermes, xvii. 148. 

Be Corona, p. 308, small fragment. Third century. Grenfell 
and Hunt, 0. P. i. 25, with facs. 

Epistle III, wanting end. First century b. c. Brit. Mus. Pap. 
cxxxiii. Kenyon, Class. Texts, p. 56, with specimen facs. 

Be Falsa Legatione, § 10, imperfect. First or second century. 
Bodl. MS. Gr. class, f. 46 (P). Grenfell and Hunt, 
G. P. ii. 9. 

Contra Leptinem, §§ 84-91, with lacunas. First or second 
century. Berl. Pap. 5879. Wilcken, Tafeln, i. 

In Meidiam, §§ 41, 42. Fourth or fifth century. In 
possession of Mr. F. Cope Whitehouse. Proceedings of 
Society of Biblical Archaeology, xv. 86, with facs. 

In Meidiam : hypothesis and part of commentary. Late 
first century. Brit. Mus. Pap. cxxxi verso. Kenyon, 
'AdrjvaiMv UoXireLa ed. 3, appendix i, with facs. 

In Meidiam, lexicon to : fragment. Eainer Coll. Men- 
tioned by Karabacek, Mitth. Ersh. Rainer, i. 51 (1886). 


Demosthenes {continued) : — 

OJynthiacs II, fragments from §§ lo, 15. First-second 
century. In library of Eossall School. Kenyon, Class. 
Bev. vi. 430 ; facs. above, pi. XVI. 

Third Philijjpic, fragment. Egypt Exploration Fund. Hogarth 
and Grenfell, E. E. F. Archaeological Beport, 1895-6, p. 17. 

Contra PJiormionem, §§ 5-7, imperfect. Second century. Bodl. 
MS. Gr. class, f. 47 (P). Grenfell and Hunt, G. P. ii. 10. 

UpooL/jita ArjixrjyoptKa, §§ 26-29, pai'ts of seven columns. First 
or second century. Grenfell and Hunt, 0. P. i. 26, -with 


Dialectical Treatise, with quotations from Homer, Sappho. 
Alcman, Anacreon, Ibycus, Thespis, Timotheus, Euripides, 
&c. : fourteen columns, imperfect. Before b. c. 160. Louvre 
Pap. 2. Letronne, Notices et Extraits, xviii. 77, with facs. 
in atlas, pi. xi. 

Dioscorides; ten fragments in a chemical papyrus. Leyden 
Pap. X. Leemans, Cat. of Leyden Papyri, ii. 205. 

Drama : — 

Anonymous : fragment, apparently of an * Iphigenia,' begin- 
nings of seventeen lines. Third centuiy b. c. Brit. Mus. 
Pap. ccccLxxxvi 1). Mahaffy, P. P. i. 3 {2), with facs. 

— parts of eighteen lines, apparently comedy. Third 
century b. c. Brit. Mus. Pap. occclxxxvii a. lb. i. 4 (i), 
with facs. 

— parts of twenty-eight lines, apparently tragedy. Third 
century b. c. Brit. Mus. Pap. occclxxxvii h. lb. i. 4 (2), 
with facs. 

— parts of twenty lines. Third century b. c. Brit. Mus. 
Pap. Dxc. Mahaffy, P. P. ii. 49 c. 

— two small fragments. Third century b. c. Brit. Mus. 
Pap. Dxci a. lb. ii. 49 d. 

— two small fragments. Tliird century b. c. Brit. Mus. Pap. 
DCi-xxxviii a. Grenfell and Hunt, G. P. ii. i a, with facs. 



Drama {continued) : — 

Anonymous : very small fragment. Third century b. c. 
Brit. Mus. Pap. dclxxxviii b. Ih. ii. i h, with facs. 

— four fragments. Third century b. c. Brit. Mus. Pap. 
Dcxc. Ih. ii. 6 a. 

— fragment. Third century b. c. Brit. Mus. Pap. dcxci a. 
lb. ii. 6 b. 

— two small fragments. Third century b. c. Brit. Mus. 
Pap. DCXCI b. lb. ii. 6 c, with facs. 

— two small fragments. Third century b. c. Brit. Mus. 
Pap. Dcxciv. lb. ii. 8 b. 

— fragment, fourteen lines. Before b. c. 161. Pap. Didot. 
Weil, Monuments Grecs, with facs. 

— fragment of comedy, parts of fifty lines. First or second 
century. Grenfell and Hunt, 0. P. i. 11. 

— fragment. Second century. Brit. Mus. Pap. cccclxxxiv d. 

— fragment of comedy, twenty lines, nine nearly perfect. 
Second or third century. Grenfell and Hunt, 0. P. i. 10. 

— parts of three or four columns, on subject of Jason. 
Second or third century. Brit. Mus. Pap. clxxxvi verso. 
Cat. of Additions, 1894. 

— fragment. Late third century. Brit. Mus. Pap. dcxcv a. 
Grenfell and Hunt, G. P. ii. 12, with facs. 

— fragment, ten lines. Berl. Pap. Blass, Zeitschr. f. dg. 
Sprache, 188 1. 

Elegiac Poetry : — 

Anonymous : fragment, parts of twenty-four lines. Third cen- 
tury b. c. Brit. Mus. Pap. dlxxxix. Mahaffy, P. P. ii. 49 a. 

— fragment, parts of eighteen lines. Second century. 
Grenfell and Hunt, 0. P. i. 14. 

Epic Poetry : — 

Anonymous : two small fragments. Third century b. c. 
Bodl. MS. Gr. class, f. 45 (P). Grenfell and Hunt, G. P. ii. 5- 


Epic Poetry {contmited) : — 
Anonymous: fragment. Second century. Brit. Mus. Pap. 
ccccLxxxiv e. Unpublished. 

— considerable fragments. Second-third century. Brit. Mus. 
Pap. ccLxxiii. Cat. of Additions, 1894. 

— fragment, fifty-four lines. Fourth century (?). At Lim- 
erick (?). Bp. Graves, Hermathena, 1885, with facs. 

— quotation, four lines. Fourth-fifth century. Paris, Bibl. 
Nat: Wilcken, Sitsungsbericlite der Bert. Akad., 1887. 

— two fragments, on subject of Phineus. Eainer Coll. 
Mentioned in Phil. Ameig. xiv. 477. 

— fragment, in book form. Fourth century. Berl. Pap. 
5003. Stern, Zeitschr. f. dg. S^racJte, 1881 ; facs. in 
Wilcken, Tafeln, v. 

Epicharmus ; fragment, parts of four lines, from an antho- 
logy. Third century b. c. Brit. Mus. Pap. cccclxxxvi a. 
Mahaffy, P. P. i. 3 (0. with facs. 

— '08vo-o-€vs ttvro/xoAos, fragment, ten lines with commentary. 
First or second century. Eainer Pap. 8023 (F. 537, with facs.), 

Epicurus, various works among Herculaneum papyri. First 
century B.C. Cf. Scott, Fragmcnta Hemdanensia. 

Epigram, anonymous, on conquest of Egypt by Augustus. 
Early first century. Brit. Mus. Pap. cclvi (2). Cat. of 
Additions, 1894 ; Rev. de Phil. xix. 177. 

Epigrams ; collection by unknown authors, fragment. Third 
century b. c. Bodl. MS. Gr. class, e. 33 (P). Mahaffy, 
P. P. ii. 49 1), with facs. 

— fragment of a collection. Thiid centuiy, Grenfell and 
Hunt, 0. P. i. 15. 

Euclid, ii. 5. Third or fourth century. Grenfell and Hunt. 
O.P. i. 29. 

Eudoxus, astronomical treatise. Early second century b. c. 
Louvre Pap. i. Brunet de Presle, Notices et Extraits, xviii. 
25, with facs. in atlas, pi. i-x. 



Euripides : — 

Antiope, iviigmeni from end of play, 123 lines. Third cen- 
tiny B. c. Brit. Mus. Pap. cccclxxxv. Mahaffy, P. P. i. i, 
with facs. 

— parts of three lines from an anthology. Third century 
B. c. Brit. Mus. Pap. cccclxxxvi a. Mahaffy, P. P. i. 
3(1), with facs. 

Medea, 11. 5-12. Before b. c. 161. Pap. Didot. Weil, 
Monuments Grecs, with facs. 

Orestes, 11. 339-343, with musical notes. Circ. a. d. i. 
Eainer Pap. 8029 (F. 531, with facs.). Wessely, 3Iitth. 
V. 65. 

— 11. 1 06 2- 1 090. Circ. second century. Geneva Pap. 
Nicole, Bev. de Phil. xix. 105. 

[Euripides], PJtesus, 11. 48-96, in book form. Fourth-fifth cen- 
tury. Paris, Bibl. Nat. Wilcken, Sitzungsl). d. Berl. AJcad. 


Euripides, Temenides (?), forty-four lines. Before b. c. 161. 
Pap. Didot. Weil, Monuments Grecs, with facs. 

Grammarian, anonymous ; fragment. First century. Eainer 
Coll. Mentioned by Karabacek, Mitth. i. 51. 

Gregory of Nyssa, Life of 3Ioses ; extracts. Fifth century (?). 
Berl. Pap. Blass, Zeitsclir. f. (ig. Spraclie, 1880. 

Hermas, Pastor, Mand. xi. 9, with an additional passage. 
Third-fourth century. Grenfell and Hunt, 0. P. i. 5 
{cf V. Bartlet, Athencetim, Oct. 8, 1898). 

— Sim. ii. 7-10, iv. 2-5. Third century. Berl. Pap. 5513. 
Wilcken, Tafeln, iii, Diels and Harnack, Sitzungsl). d. Berl. 
Akad. 1 89 1. 

Herodas, Mimes ; incomjilete. First-second century. Brit. 
Mus. Pap. cxxxv. Kenyon, Class. Texts, with specimen 
facs. ; complete facs. in separate vol. Later editions liy 
Rutherford, Biicheler, Crusius, Meister. 


Herodotus : — 

i. 76. Second or third century. Grenfell and Hunt, 0. P. 
i. 19 verso. 

— 105-6. Third century. Grenfell and Hunt, 0. P. i. 18. 

Hesiod : — 

WorTis and Days, 11. 111-118, 153-161, 174-182, 210-221, 
with four additional verses; in book form. Fifth century. 
Geneva Pap. Nicole, Bcv. de Phil. xii. 113. 

— 11. 251-266, 283-296, 313-329, 346-361, 686-709, 718- 
740, 748-812, 817-828; Shield of Heracles, 11. 5-30, 
434-440, 465-470 ; in book form. Circ. a. d. 400. 
Kainer Coll. Wessely, Mitth. i. 73. 

Theogonia, 11. 75-145, in book form. Fourth-fifth century. 
Paris, Bibl. Nat. Wilcken, Sitzungsh. d. Berl. Aliad., 1887. 

— 11. 210-238, 260-270. Fourth-fifth century. Brit. Mus. 
Pap. CLix. Kenyon, Bev. dc Phil. xvi. 181. 

Hesiod (?), Eoeae (?) ; portions of six lines. Third century 
B. c. Brit. Mus. Pap. cccclxxxvi c. Mahaffy, P. P. i. 
3 (3), with facs. 

History : — 

Anonymous, fragments of collection of vo/xtyu,a (SaplSapLKo.. 
Third century b. c. Brit. Mus. Pap. cccclxxxix. MahafPy, 
P. P. i. 9, with facs. 

— on Spartan training (? a AaKeSaLfiovtwv IloXire/a) ; fragment. 
Second centuiy. Brit. Mus, Pap. clxxxvii. Kenyon, 
Bev. de Phil. xxi. i. 

Homer : — 

Hiad i. 37-54, 65-67, 207-229 ; on the verso of accounts. 
Third century (?). Brit. Mus. Pap. cxxix. Kenyon, 
Class. Texts, p. 80. 

— i. 44-60. Circ. second century (?j. Geneva Pap. Nicole, 
Bev. de Phil, xviii. 103. 

— i. 129-150. Second century. Brit. Mus. Fap. cclxxii. 
Cat. of Additions, 1894. 


Homer {continued) : — 

Iliad i. 273-362. Second century, Gizeh Museum (from 
Egypt Exjiloration Fund). Unpublished (cf. Hogarth and 
Grenfell, E. E. F. ArcJiacological Beport, 1895-6, p. 16). 

— i. 298-333. First or second century. Bodl. MS. Gr. 
class, e. 58 (f). Unpublished, 

— i. 506, 507; ii, 1-6, 45-49, m-nS, i55-i57, 200-205, 
223-228, 245-252, 289-292, 331-337, 345-382, 391-404, 
411-422, 433-446, 454-470, 472-486, 488-492, 494-510, 
516-531, 538-560, 562-598, 601-621, 624-686, 692-731, 
735-753, 755-841, 843-877; with few scholia and 
Aristarchean symbols. Second century. Bodl. MS, Gr. 
class, a, I (P). Petrie, Haivara, p. 24 ; specimen facs. 
there and above, j)l. XX. 

— fragments of Iliad i, ii, iv, viii, xi, xvii, Kainer Coll. 
Mentioned in PMl. Anzeig. xvi. 414, 477. 

— ii. 1 01 — iv. 40 (omitting ii. 494-877), in book form. 
Third century. Brit. Mus, Pap, cxxvi. Kenyon, Class. 
Texts, p. 81, with specimen facs. 

— ii. 730-828, with an additional line after 1. 798. Second 
century. Grenfell and Hunt, 0. P. i. 20, with partial facs, 

— ii, 745-764. First or second century, Grenfell and 
Hunt, O.P.I 21. 

— iii. 317-337, 342-372; iv. 1-28, 56-69, 74-79, 11^-^50, 
159-192, 198-201, 208-245, 256-293, 303-345, 352-544; 
on the verso of tax-register. Late first century. Brit. Mus. 
Pap. cxxxvi. Kenyon, Class. Texts, p. 93, with specimen 

— iv. 82-95. First-second century 1^?;. Geneva Paj). Nicole, 
Bev. de Fhil. xviii. 103. 

— iv. 1 09-1 13. Thix-d century B.C. Brit. Mus, Pap. 
DCLxxxix l). Grenfell and Hunt, G. P. ii. 3, with facs. 

— iv. 191-219, omitting 196, 197, 215. Gizeh Museum. 
.Sayce, Academy, May 12, 1894. 

— V. 731-734, 815-818, 846-850. Second century (?;. 
Brit. Mus. Pap, cxxvii (i), Kenyon, Class. Texts, p, 98. 


Homer (continued) : — 

Iliad vi. 1-39. First-second century. Louvre Pap. s*^'". 
Longperier, Notices et Extraits, xviii. 120, with facs. in 
atlas, pi. xlix, 

— vi. 90-100, 1 1 9-1 25. Second century (?). Brit. Mus. 
Pap. cxxvri (2). Kenyon, Class. Texts, p. 98. 

— vi. 327-353 ; on verso of business document. First- 
second century (?). Geneva Pap. Nicole, Rev. de Phil. 
xviii. 104. 

— viii. 64-75, 96-116. First or second century. Bodl. 
MS. Gr. class, d. 20 (P). Grenfell, G. P. i. 2. 

— viii, a fragment. Second century b. c. Egypt Exploration 
Fund. Hogarth and Grenfell, E. E. F. Archaeological 
Report, 1895-6, p. 17. 

— viii. 217-219 (?), 249-253, with two additional lines. 
Third century b, c. Brit. Mus. Pap. dclxxxix a. Grenfell 
and Hunt, G. P. ii. 2. 

— viii. 433-447. Berl. Mus. Pap. 6845. Verzeichniss der 
dgypt. Altert. p. 370. 

— xi. 502-537, with five additional lines. Third century b. c. 
Brit. Mus. Pap. cccclxxxvi d. Mahaffy, P. P. i. 3 (4), 
with facs. 

— xi. 788 — xii. 9, with thirteen additional lines. Second 
century b. c. Geneva Pap. Nicole, Rev. de Phil, xviii. 
104: facs. in Sitzungsh. d. Berl. AMd. 1894. 

— xii. 178-198. Third century. Bodl. MS. Gr. class, e. 
21 (P). Grenfell, G. P. i. 4. 

— xiii. 26-47, 107-111, 149-173. First or second century. 
Louvre Pap. 3. Brunet de Presle, Notices et Extraits, 
xviii. 109, with facs. in atlas, pi. xii. 

— xiii. 143-150. Second century. Vatican Pap. ; part of 
same MS, as the preceding. Comptes-rcndus de VAcad. 
des Inscr. 1893. 


Homer {continued) : — 

JZ/a(Z xiii. i-io, 38-50, 73-87, 149-437, 456-653, 658-674, 
740-747; xiv. 120-293,322-394,397-522. First century. 
Brit. Mus. Pap. dccxxxii. Hunt, Journal of Philolor/>/, 
xxvi. 25. 

— xvii. 102-115, 142-152. Second centurj-. Rainer Paj). 
8027 (F. 533). 

— xviii. 1-2 18, 311-617. First century. Brit. Mus. Paj). 
cvii. Thompson and Warner, Cat. of Anc. 31 SS. ; facs. 
in Pal. Soc. ii. 64. 

— xviii. 1-22, 29-33, 77-92, 98-121, 125-136, 152-161, 
168-175, 227-230, 273-275, 279-288, 320-349, 359-371, 
387-394, 398-410, 412-425, 442-450, 455-465, 467-477, 
479-492, 5oi-5'8, 534-543, 563-575, 578-617. Second 
century ('?). Brit. Mus. Pap. cxxvii (3). Kenyon, Class. 
Texts, p. 98. 

— xviii. 475-499, 518-535, 544-561. Second century. 
Louvre Pap. 3^^'". Longperier, Notices et Exiraits, xviii. 
1 1 4, with facs. in atlas, pi. xlix. 

— XX. 36-110. Second century. Gizeh Museum (from 
Egypt Exploration Fund). Unpublished {cf. Hogarth 
and Grenfell, E. E. F. Archaeological Peport, 1895-6, p. 1 7). 

— xxi. 387-399, 607-611; xxii. 33-38, 48-55, 133-135, 
151-155, T60-262, 340-344; xxiii. 159-166, 195-200, 
224-229 ; with additional lines. Third century b. c. 
Bodl. MS. Gr. class, b. 3 (P). Grenfell and Hunt, G. P. 
ii. 4, with specimen facs. 

— xxi. 544-609 ; xxii. 390-435. Berl. Pap. Blass, Zeitschr. 
f. ag. Sp'aclie, 1880. 

— xxiii. 1-79, 402-633, 638-897; xxiv. 1-83, 100-144, 
150-243, 248-282, 337-341, 344-351, 382-387, 402-479, 
490-520, 536-548, 559-579, 596-611, 631-657, 671-729, 
737-743, 754-759 ; with a few scholia and Aristarchean 
symbols. First century b. c. Brit. Mus. Pap. cxxviii. 
Kenyon, Class. Texts, p. 100, with specimen facs., and 
Journal of Philology, xxi. 296. 


Homer {continued) : — 

Iliad xxiy. 127-804. Second century. Brit. Mus. Pap. cxiv. 
Thompson and Warner, Cat. of Anc. 3ISS., with facs. 

Odysseij iii. 267-278, 283-294, 319-335, 352-366, 389-497, 
with scholia. Early first century. Brit. Mus. Pap. cclxxi, 
with some small fragments in Eainer Coll. Kenyon, 
Journal of Philology, xxii ; specimen facs. in Pal. Soc. ii. 
182 ; Wessely, MittJi. vi. i, with hand-made facs. of 
Vienna fragments. 

— iii. 364-375, 384-402. Geneva Pap. Nicole, Rev. de 
Phil, xviii. loi. 

— V. 346-353. Third centuiy. Bodl. MS. Gr. class, g. 
7 (P). Grenfell, G. P. i. 3. 

— xiv. 15-24, 36-60, 71-86, 374-376, 378-381, 407-409, 
430-441. Berl. Pap. 154 a. Landwehr, Philologus, xliv. 

— XV, fragment. Gizeh Museum. Sayce, Academy, May 12, 

Homer : Lexicon to Uiad i ; fragment. Fifth century. Berl. 
Pap. Wilcken, Sitzungsh. d. Berl. Ahad. 1887. 

Homer : Commentary and paraphrase to Uiad i ; fragment. 
Third-fourth century. Paris, Bibl. Nat. Wilcken, ib. 

Homer : Paraphrase of Iliad iv ; fragment. Eainer Pap. 
Mentioned in Phil. Anzeig. xiv. 44. 
See also 5. v. Apollonius. 

Hyperides : — 

In Athenogenem ; impei'fect. Second century b. c. Louvre 
Pap. Ed. Eevillout, with facs. ; subsequent editions 
by Weil, Herwerden, Blass, Kenyon, Vogt. 

In Demosthenem, fragments ; Pro Lycophrone, impeifect ; 
Pro Euxenippo. Late first century. Brit. Mus. Papp. 
cviii, cxv (with fragments in Paris and at Rossall School). 
Ed. Babington, with hand-made facs. ; other editions by 
Sauppe, Boeckh, Schneidewin, Caesar, Cobet, Linder, 
Comparetti, Blass. 


Hyperides {continued j : — 

Funeral Oration, imperfect. Second century. Brit. Mus. 
Pap. xcviii verso. Ed. Babington, with hand-made facs. ; 
subsequent editions by Sauppe, Kayser, Tell, Cobet, 
Comparetti, Deheque, Caffiaux, Blass. 

In PJiilippidem, imperfect. First century e. c. Brit. Mus, 
Pap. cxxxiv. Ed. Kenyon, Class. Texts, with specimen facs. ; 
subsequent editions by Weil, Herwerden, Blass, Kenyon. 

Isocrates : 

De Antidosi, §§ 83, 87, small fragment. First or second 
century. Grenfell and Hunt, 0. P. i. 27. 

Contra Nicoclem, §§ 1-30. Third century. Marseilles Pap. 
Schoene, Melanges Graux, 481, with specimen facs. 

— §§ 2-4, fragment. Fourth century. Eainer Pap. 8029 
(F. 532). Wessely, Mittli. iv. 136 (1888). 

De Pace, §§ 1-6 1 fragmentary, rest nearly perfect. First- 
second century. Brit. Mus. Pap. cxxxii. Kenyon, Class. 
Texts, p. 63, with specimen facs. 

Pliilippus, §§ 1 1 4-1 1 7. First-second centuiy. Rainer Pap. 

Wessely, Ilitth. ii. 74 (1887). 

Isocrates : criticism of the Eragoras, anonymous. First- 
second century. Eainer Pap. Wessely, Mitlli. ii. 79 

Literary Criticism : — 

Anonymous, on the names of the gods ; fragment. Second- 
third century. Berl. Pap. 1970. Wilcken, Tafcln, ii. 

— life of Aesop. One leaf of a book. Sixth century (?). 
At St. Petersburg. Weil, Bev. de Phil. ix. 1 9. 

Lives of Saints, anonymous ; lives of SS. Abraham and 
Theodora, fragments. Louvre Papp. 7404-8 bis. Wessely, 
Wien. Stud. 1889. 

Lyric Poetry : — • 

Anonymous ; two small fragments. Third century e. c. 
Brit. Mus. Pap. dcxciii. Grenfell and Hunt, G. P. ii. 8 a. 


Lyric Poetry {continued) : — 

Anonymous ; fragment of Ittivlklov, twenty lines. Louvre 
Pap. Egger, Comptes-rendus de V Acad, des Inscr. 1877. 

— portions of three columns. Second century. Egypt 
Exploration Fund. Hogarth and Grenfell, E. E. F. 
Archaeological Beport, 1895-6, p. 16. 

Mathematics, anonymous ; fragment on mensuration of land. 
First century. Ayer Papyrus. E. J. Goodspeed, Ameri- 
can Journal of Philology, xix. 25, with facs. 

Medicine : — 

Anonymous ; treatise on diseases, with extracts from Menon's 
latrica. First-second century. Brit. Mus. Pap. cxxxvii. 
Ed. Dials, Supplementum Atistotelicum, iii. i, with speci- 
men facs. (1893). 

— fragment, on dentistry &c. First-second century. Brit. 
Mus. Pap. CLV. Cat. of Additions, 1894. 

Menander, Fewpyos, portions of one leaf, about eighty lines. 

Fourth-fifth century (?). Ed. Nicole (1897), Grenfell 

and Hunt (1898). 
Mime, anonymous, in rhythmical prose or lyrical verse ; 

imperfect. Third century b. c. Brit. Mus. Pap. dcv 

verso. Grenfell, G. P. i. i, wdth facs. 
Oratory : — 

Anonymous ; fragment. First century. In Paris (?). 

Egger, Mem. d'Hist. anc. 175 (1863). 

— remains of three rhetorical exercises, mutilated. Late 
first century. Brit. Mus. Pap. cclvi verso. Cat. of 
Additions, 1894 ; Kenyon, Melanges Weil, p. 243 (1898). 

— two columns of a private oration. Second century. 
Egypt Exploration Fund. Hogarth and Grenfell, E. E. F. 
Archaeological Beport, 1895-6, p. 16. 

— fragments. In Paris (?). Late first or second centuiy. 
Egger, Mem. de I'Acad. des Inscr. xxvi (1870), with facs. 

Pherecydes, HevTeixvxos : fragment. Third century. Bodl. 
MS. Gr. class, f. 48 (P). Grenfell and Hunt, G. P. ii. 11, 
with facs. 

APPENDIX 11 145 

Philo, Tt's 6 Tiav Oeiwv KXr)pov6fji.o? and TTcpi yci'eVews "AySeX : in book 

form. Sixth century (? third). Gizeh Museum. Ed. 

Seheil, Hem. de la Mission Arch. Frmigaise au Caire, torn. 

9 (1893), with specimen facs. 
Philodemus, various works among Herculaneum papyri. Cf. 

Scott, Fragmenta Herculanensia. 

Philosophy : — 

Anonymous ; fragment. Third century b. c. Brit. Mus. 
Pap. Dcxcii. Grenfell and Hunt, G. P. ii. 7 a. 

— five small fragments. Third century b. c. Bodl. MS. 

Gr. class, e. 63 (P). lb. 7 b. 

— fragment. Third century b. c. Brit. Mus. Pap. dxci b. 
Mahaffy, P. P. ii. 49 e. 

— several rolls among Herculaneum papyri. First cen- 
tury B.C. Cf. Scott, Fragmenta Herculanensia. 

— several fragments of a treatise, apparently philosophical, 
on verso of land-registei-. Second century. Brit. Mus. 
Pap. Dccxxxiv. 

— life of the philosoj^her Secundus. Second century (?). At 
St. Petersburg. Tischendorf, Notitia Editionis Codicis 
Sinaitici, p. 69 (i860), 

— fragment on ethics. Second or third century. Brit. 
Mus. Pap. CLxxxiv. Cat. of Additions, 1894. 

— considerable fragments on ethics. Third century (?). 
Brit. Mus. Pap. cclxxv. Ibid. 

— (? Aristotle), fragment on aesthetics. Eainer Coll. Men- 
tioned in Phil Anzeig. xiv. 414; cf. Gomperz, Mitth. 
i. 84 (1887). 

Plato : — 

Gorgias, parts of pp. 504 B-E, 505 A, in book form. Third 

century. Eainer Coll. Wessely, Mitth. ii. 76 (1887). 

Laches, small fragments of pp. 181 B-i 82 A. Second century. 

Brit. Mus. Pap. clxxxvii verso. Cat. of Additions, 1894. 

— pp. 190 B-192 A, with lacunas. Third century b. c. 
Bodl. MS. Gr. class, d. 22, 23 (P). Mahaffy, P. P. ii. 50, 
with facs. 



Plato {continued) : — 

Latvs, ix. pp. 862-3. Third century (before a. d. 295). 
Grenfell and Hunt, 0. P. i. 23, with facs. 

Phaedo, pp. 67 E-69 A, 79 C-81 J), 82 A~84 B, with 
lacunas. Third century b. c. Brit. Mus. Pap. cccclxxxviii. 
Mahaffy, P. P. i. 5-8, with facs. 
^Republic, x. pp. 607-8. Third century. Grenfell and Hunt, 
O.P. i. 24. 

Political Treatise, fragment of letter to a king of Macedon 
against the Thebans. Second or third century. Grenfell 
and Hunt, 0. P. i. 13. 

Posidippiis, two epigrams. Before b. c. 161. Pap. Didot. 
Weil, Monuments Grecs, with facs. 

Romance : — 

Anonymous, on adventures of Heracles ; fragments. Third 
century B. c. Brit. Mus. Pap. dxcii. Mahaffy, P. P. ii. 49/. 

— on adventures of Ninus ; fragments. Circ. b. c. 50-A. d. 50. 
Berl. Pap. 6926. Wilcken, Hermes, xxviii. 161 (1893). 

— on Metiochus and Parthenope ; fragment. Second cen- 
tury. Berl. Pap. 7927. Krebs, Hermes, xxx. 144 (1895). 

— with narrative of shipwreck ; fragment, on verso of 
accounts. First-second century. In Dublin ('?). Mahaffy, 
Rendiconti della R. Accad. dei Lincei, 1897, with facs. 

— considerable fragments. Second century. Brit. Mus. Pap. 
ccLxxiv. Cat. of Additions, 1894. 

Sappho (?j ; five fragmentary stanzas. Third century. Ee- 
stored by Blass, ap. Grenfell and Hunt, 0. P. i. 7, with 

Scazon Iambics, anonymous ; fragment. Third century (?). 
Brit. Mus. Pap. clv verso. Cat. of Additions, 1894. 

Science : — 

Anonyxiious, on optics. Louvre Pap. 7733. Wessely, Wien. 
Stud. xiii. 312 (1891). 


Science (continued) : — ' 

Anonymous, chemical excerpts. Third- fourth century. 
Leyden Pap. x, Leemans, Pajjijri Graeci Lugduni-Batavi, 
ii. 199 (1885). 

— on mathematics. Seventh century (?). Gizeh Museum. 
Ed. Baillet, in Mem. de la 3Iiss. Arch. Fr. au Caire, torn. 
9 {1892), with facs, 

— on astronomy ; fragment. Rainer Pap. Mentioned in 
Phil. Anzeig. xiv. 477 (1884). 

Sophocles, Oed. Tyr. 375-385, 429-441, in book form. Fifth 
century (?). Grenfell and Hunt, 0. P. i. 22. 

Theology : — 

Anonymous ; fragment, sixteen lines. Sixth-seventh cen 
tury. Brit. Mus. Pap. cxiii. 12 a. Kenyon, Cat. of Papyri, 
i. 224, with facs. (1893). 

— fragment, thirty-two lines. Sixth-seventh century. Brit. 
Mus. Pap. CXIII. 12 I). lb. p. 225, with facs. 

— fragments, forty-six lines. Sixth-seventh century. Brit. 
Mus. Pap. CXIII. 12 c. lb. p. 226, with facs. 

— fragment, twenty-six lines. Sixth-seventh century. 
Brit. Mus. Pap. cxiii. 13 a. lb. p. 227, with facs. 

- — fragment. Sixth or seventh century. Brit. Mus. Pap. 
ccccLxii. Unpublished. 

— fragment. Sixth or seventh century. Brit. Mus. Pap. 
ccccLxiv. Unpublished. 

— fragment. Berl. Pap. Blass, Zeitschr. f. dg. Sprache, 

— fragment, on verso of papyrus. Fourth century. Gren- 
fell and Hunt, 0. P. i. 4. 

Thucydides : — 

ii. 7, 8. Second or third century. Grenfell and Hunt, 0. P. 
i. 17. 

iv. 36-41. First- second century. Hunt, U. E. F. Archaeo- 
logical licport, 1896-7, J). 13, and 0. P. i. 16, with partial 

L 2 


Travels ; fragment of a description of Athens. Second century. 
Petrie, Ilawara, p. 28 (1889); cf. Wilcken, JBerl. Phil. 

WocJienschrift, Dec. 7, 1889. 

Tryphon, rexi'V ypa/x/Aart/o/, imperfect, in book form. Early 
fourth century. Brit. Mus. Pap. cxxvi verso. Kenyon, 
Class. Texts, with specimen facs. 

Xenophon : — 

Cyropaedia, v. 2. 3-3. 26, imperfect. Second century. Eainer 
Coll. Wessely, IliUh. vi. i (1897). 

Hellenica, i. 2. 2-5. 8, imperfect. Third century. Eainer 
Coll. Wessely, ih. vi. 17. 

— iii. I, parts of three columns. Second century (?). 
Grenfell and Hunt, 0. P. i. 28. 

Memorabilia, i. 3. 15, 4. 1-3. Third-fourth century. 
Brit. Mus. Pap. dcxcv h. Grenfell and Hunt, G.P.ii. 13. 



(Arranged chronologically under the respective countries. 
Isolated publications in periodicals are not included ; they 
will mostly be found in Hermes, Philologus, and the Revue de 

I. Austria. 

1. Pettretini (G.), Papiri greco-egizi ed altri greci monumenti 

delV I. It. Museo di Corte (Vienna, 1826), with three plates. 

2. Peyron (A.), Papiri greco-egizi di Zoide delV I. B. 3Itiseo 

di Vienna, with two plates (Turin, 1828); a revision 
of some of Pettretini's texts. 

3. Wessely (K.), articles in Wiener Studien, iii-v, vii, xi 

(Vienna, 1 881-1889). 

4. Wessely (K.), articles in MittJieilungen axis der Sammhwg 

der Papyrus Erzherzog Bainer (Vienna, 1887-1897). 

5. Wessely (K.), Papyrus Erzlierzog Bainer: Filhrer durcli die 

Ausstellung (Vienna, 1894) ; edited by J. Karabacek, the 
Greek section by Wessely, with nine facsimiles of Greek 

6. Wessely (K.), Corpus Papyrorum Baineri, vol, i. Griechi.sche 

Texte (Vienna, 1895). 

II. France. 

7. Brunet de Presle (W.), Notices et Tcxtcs des Papyrus Grecs 

du Musee de Louvre et de la BihliotMque Imperiale (in 
Notices et Extraits des Ilanuscrits de la BihliotMque Imperiale 
et autres Bihliotheques, vol. xviii), with atlas of fifty-two 
plates (1865). 


8. Wessely (K.), Lettres a M. Eevillout (in Bevue Egy})tologiqiie, 


9. Wessely (K.), article in Wiener Studien, viii (1886). 

10. Wessely (K.), Griecliische Zauherpapyrus von Paris und 

London {DenJ:schriftcn der Jcais. Alademie der Wissen- 
scliaften, Vienna, 1888). 

1 1 . Wessely (K.), Die Pariser Papyri dcs Fundcs von El-Faijum 

iih. 1889). 

12. Wessely (K.), Zu den griechisehen Papyri des Louvre und 

der BihliotMque Nationale (in Jahresbcricht des Jc. k. 
Staatsgymnasiians Hernals, two parts, Vienna, 1889, 

13. WiTKowsKi (S.\ Prodromus grammaticae papyrorum Grae- 

canim aetatis Lagidarum (Cracow, 1897) ; includes 
revision of no. 7. 

III. Great Britain and Ireland. 

1 4. Forshall (J.), Greeli Papyri in the British Museum (London, 

15.PEYRON (B.), Papiri Greci del Museo Britannico di Londra 
e dclla Bihliotheca Fafimwa (Turin, 1841); Forshall's texts 
re-edited with commentary. 

16. Wessely (K.), articles in Wiener Studien, viii, ix, xii 

(Vienna, 1886, 1887, 1890). 

17. Wessely (K.), Griechische Zatdjerpapynis von Paris und 

London (DenJcscJiriften der k. Akademie der Wissenschaften, 
Vienna, 1888) ; = no. 10 above. 

18. Wessely (K.), Neue Griechische Zauberpapyri {ib. 1893). 

19. Mahaffy (J. P.), The Flinders Petrie Papyri; Part I, with 

thirty plates (Dublin, 1891); Part II, with eighteen 
plates (Dublin, 1893); appendix, with three plates 
(Dublin, 1894). 

20. Kenyon (F. G.), Greek Papyri in the British Mtiscmn, 

Catalogue ivith Texts; vol. i, with atlas of 150 plates 
(London, 1893) : vol. ii, with atlas of 123 plates (London, 


21. Kenyon (F. Gr.), Catalogue of Additions to the Department of 

Manuscripts in the British Museum, 1 888-1 893, pp. 
390-450 (descriptions of papyri cxxi-cccclviii in the 
British Museum). 

22. Grenfell (B. p.), Greek Papyri from Apollonopolis {Journal 

of Philology, 1894). 

23. Grenfell (B. P.) and Mahaffy (J. P.), Revenue Laws of 

Ptolemy Pliiladelphus, with thirteen plates (Oxford, 1896). 

24. Grenfell (B. P.), An Alexandrian Erotic Fragment and 

other Greek Papyri (Greek Papyri I), with one plate 
(Oxford, 1896). 

25. Grenfell (B. P.) and Hunt (A. S), New Classical Fragments 

and other Greek and Latin Papyri (Greek Papyri II), 
with five plates (Oxford, 1897). 

26. Grenfell (B. P.) and Hunt (A. S.), TJie Oxyrhynchus 

Papyri, Part I, with eight plates (Egypt Exploration 
Fund, London, 1898). 


IV. Germany. 

Boeckh (A.), Erklanmg einer dgyptischcn TJrkunde auf Papyrus 
{Ahhandlungen der Berl. Akademie, Berlin, 1821). 

28. Letronne (J. A.), Catalogue des antiquites decouvertes en 

EgypteparM. Passalacqua, with a facsimile (Paris, 1826). 

29. Schmidt (W. A.), Die griechischen Papyrusurkunden der 

kdniglicJien BiUiothck zu Berlin, with two facsimiles 
(Berlin, 1842). 

30. Parthey (G. F. C.), Zaulerpapyri {Ahhandlungen der konigl. 

A kademie zu Berlin, 1865). 

31. Parthey (G. F. C), Die griechische Papyrusfragmente der 

Leipziger UniversitdtshiUiothek {Sitzungsherichtc der konigl. 
Akademie zu Berlin, 1865). 

32. Wessely (K.), Die griechischen Papyri Sachsens {Berichte der 
konigl. Sachs. Gesellschaft der Wissenschaften, 1885). 

Magirus (K.) and Wessely (K.), Griechische Papyri im 
agyptischen Museum in Berlin (in Wiener Studien, Vienna, 



34. WiLCKEN (U.), ArsinoitischeSteuerprofessionenundverwandte 

Urlamdcn, with four plates [Sitsiingsbericlite der JconigL 
AJcademie zu Berlin, 1883). 

35. WiLCKEN (U.), ActenstiicJie aiis der lionlgliclien Banlc zu Thebcn 

in den Mitscen von Berlin, London, Paris {AhJiandlungen 
derJconigl. Akademie zu Berlin, 1886). 

36. WiLCKEN (U.), articles in Hermes, xix, xx, xxi, xxii, xxiii, 

xxvii, xxviii, xxix, xxx (1884-1895), and Philologus, liii 

37. WiLCKEN (U.), Tafeln zur alteren griecliisclien Paldograplue 

(Leipzig and Berlin, 1891) ; twenty plates. 

38. WiLCKEN (U.), Krebs (F.), and Viereck (P.), GriecMsclie 

UrA-unden aus den Museen in Berlin, vol. i, with tliree 
facsimiles (1892-1895), vol. ii, with facs. of Latin papyrus 
(189 5- 1898), vol. iii in progress. 

V. Holland. 

39. Eeuvens (C. J. C), Lettres a M. Letronne sur les Papyrus 

hilingues et grecs . . . du Musee d'Antiquites de VUniversite 
de Leide (Leyden, 1830). 

40. Leemans (C), Papyri &raeci Musei Antiquarii Puhlici 

Lugduni-Batavi, vol. i, with six plates (1843); vol. ii, 
with four plates (1885). 

VI. Italy. 

41. ScHow (N.), Charta 'papiyracea graece scripta Mtfsei Borgiani 

Velitris, with six plates (Eome, 1788). 

42. FuRiA (F. del), Lllustrazione di un papiro Greco, clie si 

conserve presso il cli. sig. Luigi Lamhruschini (in Furia's 
Collezione, vol. 17, Florence, 181 3). 

43. Peyrok (A.), Papyri Gracci Begii Taurinensis Musei Aegyptii, 

with six plates (Turin, 1826). 

44. Mai (A.), Aiictores Classici, vol. iv. 445, v. 352, 356, 601, 

602, 603 (Eome, 1831, 1832) ; texts of six papyri in the 


45. Peyron (B.), Papiri Greci .... della Bibliotheca VaUcana 

(Turin, 1841) ; = no. 15 above. The Vatican papyri are 
four of those given by Mai, with commentary. 

VII. Switzerland. 

46. Nicole (J.), articles in Eevue de PMMogie, xvii, xx (1893, 

1896), Pevue ArcJteologique, xxiv, xxv (1893, 1894), Revue 
des Ehides Grecques, 1895. 

47. Nicole (J.), Les Pcvpyrus de Geneve, vol. i. fasc. i (1896). 



In the following list, A =the Brit. Mus. papyrus of Aristotle's 'A9r]va'tm' 

H = Herculaneum Papyrus 157-152. 
M = Medical Papyrus in the British Museum. 

Tables of abbreviations and symbols used in non-literary papyri will 
be found in the indices to the British Museum Catalogue of Papyri, 
Grenfell's Greek Papyri, vols, i and ii, and the Griechische Urkunden 
of the Berlin Museum. All except those which are the caprices of an 
individual, or as to which some doubt attaches, are included in the 
list given below. In late theological papyri the ftimiliar abbreviations 
fls, Ks, avos K.T.X. are found, as in contemporary vellum MSS. 




^ = aL (as part of a 


^ = daiv (A, M). 


/ = Icrn (A, H, M). 

CfjL = atrtOi, ahia, k.t.X. (M). 

6' = em (A). 

a = avd (A). 

/ = Kat (A). 

^ = avTos and cases (A and 

K = KUL (H). 


K = Kara (M). 

y = ydp (A, M). 

K = Kara (A). 

y = yap (H). 

A = Aoyos-(M, magical papp.) 

I^ = yiverca K.T.k. (M). 

"p.' = i,ev (A, M). 

8' =U{h). 

fx = juerd (A). 

8^ =Oici(A). 

N = vov (M). 

8" — hvvaixis and cases 


N^ = vr}S (M). 

\ = dvca (A, H, M). 

N = vcov (M). 



0' = ovv (A). 


= TI-jV (A). 

5 =1 oi/Vco (M). 


= 77] S (A). 

■n ■=■ irapd (A, M). 


= TpoTTos and cases (H). 

it' = 776/1 or TTepi (A). 


= TMV (A, M). 

7r' = TTpoj (M). 


= TiOV (H). 

U = 7rot?/T7/s- (Harris He 



= vTrdpx^i- K.T.X. (M). 

jf[ = TTporepov (M). 



= vTiep (A). 

ffl = Trpo? (H, Brit. 



= ^TTO (A). 

Pap. CCLVI verso). 


= (f)ricr[i', K.T.X. (M). 

a' = (Tvv (A). 


= Xpdvos (A, H). 


(/ = aTTO. 

-ss = dpovpa (rare). 
■/j = dpovpa (common). 
( = dpTa^i] (rare). 
[^ — dprd^t] (rare). 
— = dpTajBi] (common). 
5 = avTos and cases. 
A = geira (in magical papp.). 
f- = bpaxiJ-i] (Ptolemaic). 
( = hpaxp-ri (Roman). 
f = hpaxP'')] (Roman, less 

^ = bpaxp-v (Roman). 
L = eVos and cases. 
S = TJ'Aio? and cases (magical 

5 = Kai' (?). 
S = /cat. 
y 3= Kepdrioi'. 
t = /xerprjT?/?. 
n , El , or OfJ-a = ovopa 

(magical papp.). 
□ , Q S , □□ , or D ra = 6v6- 

fxara (magical papp.). 


Q = 

P = 

t = 

([ = 

A = 

Q: = 

t = 

r = 

F = 

^ = 

o = 

s = 


TTou-Jixa (magical papp.). 

•npayp-a (magical papp.). 


Tivpov, or loosely Ttvpov 

a-eXrjvr] and cases 

(magical papp.). 
a-KTJTTTpov (magical 

virep (Byzantine). 

Xpie, K.T.X. 

1 obol. 

2 obols. 

3 obols. 

4 obols. 

5 obols. 

2 chalchi. 
4 chalchi or 
: 6 chalchi. 





9 =90. 

M =i 

T' =: 900. 

/ ■= y!v(.Tai, or total. 

a =v- 

r = yiverai, or total. 

■> 2 

L = sign of subtraction. 


~) = remainder. 

a ^i\ 

n = remainder [ir^pUa-n). 

' A combination of the signs for ^ and ^. 

- The explanation of this sign is given by Brit. Mus. Pap. Dciv ; it is 
the letter A ( = 4) wi-itten like (as often in numerals, e. g. fo constantly = 
^\), and with a stroke above it to mark that it is a fraction {6, hence d). A S 
of the form d does not appear till the fourth century. 


(Figures in heavy type indicate that the principal description of tiie 
MS. in question is to be found on the page so indicated.) 

A, importance of, as a test letter, 


Abbreviations, 32. 

Abinnaeus, papyri of, 7, 47. 

Accentuation, 28, 

AlemaTi, Louvre fragment of, 5, 28, 

32, 82. 
Alexancbine MS. of the Bible, 121. 
ApoUonius, fragment of Homeric 

Lexicon, 87. 
Arden (J.), papyrus of Hyperides 

discovered by, 5. 
Aristarchean marks, 31, 81. 
Aristotle, 'AOrjvaicav noA.jT«ta. papyrus 

of, 6, 17, 21, 22, 32, 44,56, 91. 
Arsinoe, great find of papyri at, 6. 
Artemisia, Curse of, 28, 57. 
Ayer mathematical papyrus, 100. 

Babington (Churchill), 85, 103. 

Bacchylides, papyrus of, 7, 18, 27, 
28, 29, 30, 31, 75. 

Bankes (W. J.), papyrus of Homer 
discovered by, 5 ; and see s.v. 

Bernard (J. H.), 118. 

Bible, earliest extant papyrus frag- 
ments of, 109 ; early vellum uncial 
MSS. of, 120. 

Birt (T.), on manufacture of 
papyrus, 15. 

Blass (F.), 3, 57, 77, 86, 95, 107. 

Borchardt (L.), on length of papyrus 
rolls, 18. 

Borgia (Cardinal Stefano), 4. 

Breathings, 30. 

British Museum Papyri, not indexed 
under authors' names (mostly non- 
literary) : 
Pap. XXIV, 39. 
XLiv, ib. 
XLVi, 116. 

LXXVII, 52. 

xcvni recto, 103. 
ox, 104. 
cxix, 22. 
cxxi, 20, no. 
cxxxvii, 20. 
cxLi, 88, loi. 

CXLIII, 43. 
CLXXXVII, 90, 102. 

ccviiic, 87. 
ccx, 51. 

CCLVII-CCLIX, 18, 86. 

ccLX, 2 2, 43, 86. 

CCLXVI, 20. 

cccxxxn, 45. 

CCCLT, 46. 
CCCLIV, 16, 82. 
DCXXIII, 2 2, 53. 
DCLXXV, 2 2. 



Codices, papyrus, 24. 

Corrections, how indicated in papyri, 


Cyril, Dublin papyrus of, 1 1 8. 

Dates, how expressed in papyri, 53. 
Demosthenes, Xlpooifxia SrjfxrjyoptKa, 

among Oxj'rhynchus papyri, 76. 
Demosthenes, third Epistle, papyrus 

of, 16, 77. 
Demosthenes, De Falsa Legutione, 

vellum fragment of, 113. 
Diaeresis, marks of, 30. 
Dialectical treatise, in Louvre, 65. 
Dioscorides, Vienna MS. of, 121. 
Diphthongs, accentuation of, 29. 
Division of words at eud of line, 31. 

Egyptian and extra-Egyptian writing, 

Erotic Fragment, 28, 68. 

Eta, peculiar form of, 44. 

Eudoxus, astronomical treatise de- 
rived from, in Louvre, 56. 

Euripides, Antiope, papyrus of, 6, 
19, 27, 29, 61. 

Euripides, Oredes, Geneva fragment 
of, 100. 

Ezekiel, Bodleian frngment of, 107. 

Fayyum, discoveries of papyri in, 4, 
5, 6, 49. 

Festal Letter, fragment of, 118. 

Forshall (.T.\ editor of British Mu- 
seum ]iapyri, 5. 

Gardthausen (V.), treatment of papyri 
in his Griechische Paldographie, 

2, 15- 
Goodspeed (E. J.), 100. 
Grenfell (B. P.), 7, 36, 40, 41, 47, 59, 

60, 68, 76, 87, 91, 97, 106, 107, 


Harris (A. C), papyrus of Hyperides 
discovered by, 5 ; papyri of Homer, 
ih. ; and see s.v. Homer. 

Heracles, papyrus fragment of adven- 
tures of, 60. 
Herculaneum, papyrus rolls of, 3, 32, 

Hermas, fragment of, at Berlin, 106. 
Hermopolis, papyri discovered at, 6. 
Herodas, Mimes, papyrus of, 6, 16, 

17, 18, 22, 27, 29, 31, 94. 
Herodotus, on use of skins as writing 

material, 14. 
Hesiod, British Museum fragment of, 

Hogarth (D. G.), 7, 40, 41. 
Homer, Iliad II, Oxford papyrus of, 

Homer, Iliad II-IV, papyrus of (B. 

M. Pap. cxxvi), 25, 27, 29, 105. 
Homer, Iliad III, IV, papyrus uf 

(B. M. Pap. cxxxvi), 96. 
Homer, Iliad XI, Geneva fragment 

of, 68. 
Homer, Iliad XIII, XIV, papyrus 

of (B. M. Pap. Dccxxxii I, 97. 
Homer, Iliad XVIII, papyrus of 

(Harris Homer, B. M. Pap. cvil), 

22, 23, 28, 29, 30, 84. 
Homer, Iliad XXIII, XXIV, papy- 
rus of (B. M. Pap. cxxviii), 17, 18, 

21, 28, 29, 31, 81. 
Homer, Iliad XXIV, papyrus of 

(Bankes Homer, B. M. Pap. cxiv), 

28, 29, 30, 73, 98. 
Homer, ///«(?, fragments of, in Louvre, 

5, 99. 
Homer, Hind, Milan MS. of, on vel- 
lum, 121. 
Homer, Odyssey III, British Museuu) 

papyrus of, 16, 18, 21, 28, 29, 30. 

31, 83. 
Hunt (A. S.\ 7, 76, 91. 
Hyperides, In Athenogenem, Louvre 

papyrus of, 6. 18, 21, 27, 64, 65, 67. 
Hyperides, In Demosthenem, Fro 

Lycophrone, and Pro JEuxenippo, 

papyrus of, 5, 17, 18, 21, 22, 27, 

Hyperides, Funeral Oration, papyrus 

of, 5, 103. 



Hyperides, In Philippidem, papyrus 
of, 6, 16, 18, 21, 22, 27, 77. 

Indiction-period, dating by, 54. 

Inscriptions, bearing of, npon palaeo- 
graphy, 9. 

Isaiab, fragment of, at Vienna, 109. 

Isocrates, Be Pace, papyrus of, iS, 
21, 22, 95. 

Isocrates, In Nieoclem, Marseilles 
papyrus of, 27, 108. 

Kepara, 23. 
KuWrjua, 1 6. 

Leemans (C), editor of Ley den 

papyri, 5. 
Logia letnc, pajjyrus of, 7. 

Magical papyrus at Paris, 115. 
Mahaffy (J. P.), 35, 60, 62, 6S. 
Mariette (F. A. F.), papyrus of Alcnian 

discovered by, 5. 
Matthew, St., Oxyrhynchus fragment 

of, 109. 
Medical papyrus in British Museum, 

20, 32, 56. 
Menander, Tfwpyos, papyrus of, 7, 25. 
Minuscule hand on vellum, origin 

of, 51, 124. 
Montfaucon (B. de), 3. 

New Testament, the earliest MSS. 

of, how written, 93. 
Nicholson (E. W. B.), Bodley's 

Librarian, 87. 
Nicole (.J.), 7, 68, 100. 

'O/XfpaXu?, 23. 

Oxyrhynchus, papyri found at, 7, 47, 
49. 76, 77. 91. 100. 103- 108. 

Palaeogrnphy, uncertainty of, 79. 
Papyri, history of discoveries of, 3 ; 

chronological period covered by, 7 ; 

distinction between literary and 

non-literary, 9 ; value of the latter 

for palaeography, 12. 
I'apyi'us as writing material, early 

use of, 14; preparation for use, 

15; dimensions of sheets, 16 ff . ; 

recto and verso, 19 ; width of 

colrmns, 21 ; use of rollers, 23; its 

use in book form, 24. 
Tiapaypatpos, 27. 

Periods of papyrus-palaeography, 34. 
Peter, Gospel and Revelation of, 

vellum MS. of, 119. 
Petrie (W. M. F.), papyri discovered 

by, 6, 35, 36, 58, loi. 
Peyron (A.), editor of Turin papyri, 5. 
^atvuKrjs, 24. 

Pherecydes, fi-agnient of, 107. 
Philodemus, pajiyri of, at Herca- 

laneura, 7^. 
PMlijra, 15. 

Plato, Laches, papyrus of, 63. 
Plato, Laws, Oxyrhynchus fragment 

of, 109. 
Plato, Phaedo, papyrus of, 6, 19, 27, 

28, 29, 61. 
Pliny, on tlje preparation of papyrus, 

15, 16, 17. 
Prisse Papyrus, 14. 
Psalter, fragment of (B. M. Pap. 

ccxxx), 109. 
Psalter, portion of (B. M. Pap. 

xxxv]i\ 116. 
Punctuation, 27, 

Rainer collection of papyri, 47. 
liecto and Verso, rule of, 19. 
Revenue Papyrus, 17, 20, 35. 
Rollers, use of, 23. 

Sappho, Oxyrhynchus fragment of, 

Scapus, 18. 
Schow (N.), 4. 
Scott (W.), edition of Herculanean 

papyri, 71, 74. 
SeAi's, 21. 
Separation of words, 26. 



Serapeum at Memphis, papyri of, 4, 

^iWvPos, 2 2. 

Sinaitic M8. of the Bible, 120. 
Sloping hand, use of in third centur}', 

Socnopaei Nesus, great find of papyri 

at, 6. 
Stobart (H.), papyrus of Hyperides 

discovered by, 6. 

Tachygraphy, 33. 

Thompson (Sir E. Maunde), 57, 62, 
84, 85, 98, loi ; his Handbook of 
Greek and Latin Palaeography, 3, 

15. 24> 36, 74> "3- 

Titles, 22. 

Trvphon, papyrus of treatise attri- 
buted to, 27, 106. 

Uncial and cursive writing, dis- 
tinction between, with regard to 
papyri, 10. 

Vatican MS. of the Bible, 1 20. 
Vellum, earliest use of, 112 ; papyrus 

superseded by it, 115; results of 

the change, 122. 

Wattenbach (VV.), 3, 31. 

Wessely (K.), 57. 

Wilcken (U.), 19, 106, 124; his 

Tafeln zur dlteren griechiseJien 

FaldograpMe, 46, 47. 
Writing, early use of, 8. 

H, importance of, as a test letter, 73. 

Zechariah, papyrus of, 25, 118. 






■^rW- ".'^'laW 

'.iJjm^liiii' j'l*"